End-to-end pipeline for producing publication-ready ML/AI research papers targeting **NeurIPS, ICML, ICLR, ACL, AAAI, and COLM**. This skill covers the full research lifecycle: experiment design, execution, monitoring, analysis, paper writing, review, revision, and submission.
This is **not a linear pipeline** — it is an iterative loop. Results trigger new experiments. Reviews trigger new analysis. The agent must handle these feedback loops.
1.**Be proactive.** Deliver complete drafts, not questions. Scientists are busy — produce something concrete they can react to, then iterate.
2.**Never hallucinate citations.** AI-generated citations have ~40% error rate. Always fetch programmatically. Mark unverifiable citations as `[CITATION NEEDED]`.
3.**Paper is a story, not a collection of experiments.** Every paper needs one clear contribution stated in a single sentence. If you can't do that, the paper isn't ready.
4.**Experiments serve claims.** Every experiment must explicitly state which claim it supports. Never run experiments that don't connect to the paper's narrative.
5.**Commit early, commit often.** Every completed experiment batch, every paper draft update — commit with descriptive messages. Git log is the experiment history.
### Proactivity and Collaboration
**Default: Be proactive. Draft first, ask with the draft.**
| Confidence Level | Action |
|-----------------|--------|
| **High** (clear repo, obvious contribution) | Write full draft, deliver, iterate on feedback |
results/ # Raw experiment results (auto-generated)
tasks/ # Task/benchmark definitions
human_eval/ # Human evaluation materials (if needed)
```
### Step 0.3: Set Up Version Control
```bash
git init # if not already
git remote add origin <repo-url>
git checkout -b paper-draft # or main
```
**Git discipline**: Every completed experiment batch gets committed with a descriptive message. Example:
```
Add Monte Carlo constrained results (5 runs, Sonnet 4.6, policy memo task)
Add Haiku baseline comparison: autoreason vs refinement baselines at cheap model tier
```
### Step 0.4: Identify the Contribution
Before writing anything, articulate:
- **The What**: What is the single thing this paper contributes?
- **The Why**: What evidence supports it?
- **The So What**: Why should readers care?
> Propose to the scientist: "Based on my understanding, the main contribution is: [one sentence]. The key results show [Y]. Is this the framing you want?"
### Step 0.5: Create a TODO List
Use the `todo` tool to create a structured project plan:
```
Research Paper TODO:
- [ ] Define one-sentence contribution
- [ ] Literature review (related work + baselines)
- [ ] Design core experiments
- [ ] Run experiments
- [ ] Analyze results
- [ ] Write first draft
- [ ] Self-review (simulate reviewers)
- [ ] Revise based on review
- [ ] Submission prep
```
Update this throughout the project. It serves as the persistent state across sessions.
### Step 0.6: Estimate Compute Budget
Before running experiments, estimate total cost and time:
```
Compute Budget Checklist:
- [ ] API costs: (model price per token) × (estimated tokens per run) × (number of runs)
- [ ] GPU hours: (time per experiment) × (number of experiments) × (number of seeds)
**When budget is tight**: Run pilot experiments (1-2 seeds, subset of tasks) before committing to full sweeps. Use cheaper models for debugging pipelines, then switch to target models for final runs.
### Step 0.7: Multi-Author Coordination
Most papers have 3-10 authors. Establish workflows early:
| **Git + LaTeX** | `git` with `.gitignore` for aux files | Technical teams, need branch-based review |
| **Overleaf + Git sync** | Overleaf premium | Best of both — live collab with version history |
**Section ownership**: Assign each section to one primary author. Others comment but don't edit directly. Prevents merge conflicts and style inconsistency.
```
Author Coordination Checklist:
- [ ] Agree on section ownership (who writes what)
- [ ] Set up shared workspace (Overleaf or git repo)
**Load the `arxiv` skill** for structured paper discovery: `skill_view("arxiv")`. It provides arXiv REST API search, Semantic Scholar citation graphs, author profiles, and BibTeX generation.
Use `web_search` for broad discovery, `web_extract` for fetching specific papers:
**Recommended**: Install **Exa MCP** for real-time academic search:
```bash
claude mcp add exa -- npx -y mcp-remote "https://mcp.exa.ai/mcp"
```
### Step 1.2b: Deepen the Search (Breadth-First, Then Depth)
A flat search (one round of queries) typically misses important related work. Use an iterative **breadth-then-depth** pattern inspired by deep research pipelines:
```
Iterative Literature Search:
Round 1 (Breadth): 4-6 parallel queries covering different angles
- "[method] + [domain]"
- "[problem name] state-of-the-art 2024 2025"
- "[baseline method] comparison"
- "[alternative approach] vs [your approach]"
→ Collect papers, extract key concepts and terminology
Round 2 (Depth): Generate follow-up queries from Round 1 learnings
- New terminology discovered in Round 1 papers
- Papers cited by the most relevant Round 1 results
- Contradictory findings that need investigation
→ Collect papers, identify remaining gaps
Round 3 (Targeted): Fill specific gaps
- Missing baselines identified in Rounds 1-2
- Concurrent work (last 6 months, same problem)
- Key negative results or failed approaches
→ Stop when new queries return mostly papers you've already seen
```
**When to stop**: If a round returns >80% papers already in your collection, the search is saturated. Typically 2-3 rounds suffice. For survey papers, expect 4-5 rounds.
**For agent-based workflows**: Delegate each round's queries in parallel via `delegate_task`. Collect results, deduplicate, then generate the next round's queries from the combined learnings.
### Step 1.3: Verify Every Citation
**NEVER generate BibTeX from memory. ALWAYS fetch programmatically.**
For each citation, follow the mandatory 5-step process:
```
Citation Verification (MANDATORY per citation):
1. SEARCH → Query Semantic Scholar or Exa MCP with specific keywords
2. VERIFY → Confirm paper exists in 2+ sources (Semantic Scholar + arXiv/CrossRef)
3. RETRIEVE → Get BibTeX via DOI content negotiation (programmatically, not from memory)
4. VALIDATE → Confirm the claim you're citing actually appears in the paper
5. ADD → Add verified BibTeX to bibliography
If ANY step fails → mark as [CITATION NEEDED], inform scientist
```
```python
# Fetch BibTeX via DOI
importrequests
defdoi_to_bibtex(doi:str)->str:
response=requests.get(
f"https://doi.org/{doi}",
headers={"Accept":"application/x-bibtex"}
)
response.raise_for_status()
returnresponse.text
```
If you cannot verify a citation:
```latex
\cite{PLACEHOLDER_author2024_verify_this}% TODO: Verify this citation exists
```
**Always tell the scientist**: "I've marked [X] citations as placeholders that need verification."
See [references/citation-workflow.md](references/citation-workflow.md) for complete API documentation and the full `CitationManager` class.
### Step 1.4: Organize Related Work
Group papers by methodology, not paper-by-paper:
**Good**: "One line of work uses X's assumption [refs] whereas we use Y's assumption because..."
**Bad**: "Smith et al. introduced X. Jones et al. introduced Y. We combine both."
---
## Phase 2: Experiment Design
**Goal**: Design experiments that directly support paper claims. Every experiment must answer a specific question.
**Artifact preservation** — save all intermediate outputs:
```
results/<experiment>/
<task>/
<strategy>/
final_output.md # Final result
history.json # Full trajectory
pass_01/ # Per-iteration artifacts
version_a.md
version_b.md
critic.md
```
**Separation of concerns** — keep generation, evaluation, and visualization separate:
```
run_experiment.py # Core experiment runner
run_baselines.py # Baseline comparison
run_comparison_judge.py # Blind evaluation
analyze_results.py # Statistical analysis
make_charts.py # Visualization
```
See [references/experiment-patterns.md](references/experiment-patterns.md) for complete design patterns, cron monitoring, and error recovery.
### Step 2.5: Design Human Evaluation (If Applicable)
Many NLP, HCI, and alignment papers require human evaluation as primary or complementary evidence. Design this before running automated experiments — human eval often has longer lead times (IRB approval, annotator recruitment).
**When human evaluation is needed:**
- Automated metrics don't capture what you care about (fluency, helpfulness, safety)
- Your contribution is about human-facing qualities (readability, preference, trust)
- Reviewers at NLP venues (ACL, EMNLP) expect it for generation tasks
**Key design decisions:**
| Decision | Options | Guidance |
|----------|---------|----------|
| **Annotator type** | Expert, crowdworker, end-user | Match to what your claims require |
| **Scale** | Likert (1-5), pairwise comparison, ranking | Pairwise is more reliable than Likert for LLM outputs |
| **Sample size** | Per annotator and total items | Power analysis or minimum 100 items, 3+ annotators |
| **Agreement metric** | Cohen's kappa, Krippendorff's alpha, ICC | Krippendorff's alpha for >2 annotators; report raw agreement too |
| **Platform** | Prolific, MTurk, internal team | Prolific for quality; MTurk for scale; internal for domain expertise |
**Annotation guideline checklist:**
```
- [ ] Clear task description with examples (good AND bad)
- [ ] Decision criteria for ambiguous cases
- [ ] At least 2 worked examples per category
- [ ] Attention checks / gold standard items (10-15% of total)
- [ ] Qualification task or screening round
- [ ] Estimated time per item and fair compensation (>= local minimum wage)
- [ ] IRB/ethics review if required by your institution
```
**Reporting requirements** (reviewers check all of these):
- Number of annotators and their qualifications
- Inter-annotator agreement with specific metric and value
- Annotation interface description or screenshot (appendix)
- Total annotation time
See [references/human-evaluation.md](references/human-evaluation.md) for complete guide including statistical tests for human eval data, crowdsourcing quality control patterns, and IRB guidance.
---
## Phase 3: Experiment Execution & Monitoring
**Goal**: Run experiments reliably, monitor progress, recover from failures.
**Parallel execution**: Run independent experiments simultaneously, but be aware of API rate limits. 4+ concurrent experiments on the same API will slow each down.
### Step 3.2: Set Up Monitoring (Cron Pattern)
For long-running experiments, set up periodic status checks. The cron prompt should follow this template:
```
Monitor Prompt Template:
1. Check if process is still running: ps aux | grep <pattern>
2. Read last 30 lines of log: tail -30 <logfile>
3. Check for completed results: ls <result_dir>
4. If results exist, read and report: cat <result_file>
5. If all done, commit: git add -A && git commit -m "<descriptive message>" && git push
6. Report in structured format (tables with key metrics)
7. Answer the key analytical question for this experiment
```
**Silent mode**: If nothing has changed since the last check, respond with `[SILENT]` to suppress notification to the user. Only report when there's news.
### Step 3.3: Handle Failures
Common failure modes and recovery:
| Failure | Detection | Recovery |
|---------|-----------|----------|
| API rate limit / credit exhaustion | 402/429 errors in logs | Wait, then re-run (scripts skip completed work) |
| Process crash | PID gone, incomplete results | Re-run from last checkpoint |
| Timeout on hard problems | Process stuck, no log progress | Kill and skip, note in results |
| Wrong model ID | Errors referencing model name | Fix ID and re-run |
**Key**: Scripts should always check for existing results and skip completed work. This makes re-runs safe and efficient.
Git commits track what happened, but not the **exploration tree** — the decisions about what to try next based on what you learned. Maintain a structured experiment journal that captures this tree:
```json
// experiment_journal.jsonl — append one entry per experiment attempt
{
"id":"exp_003",
"parent":"exp_001",
"timestamp":"2025-05-10T14:30:00Z",
"hypothesis":"Adding scope constraints will fix convergence failure from exp_001",
"plan":"Re-run autoreason with max_tokens=2000 and fixed structure template",
"analysis":"Scope constraints fixed convergence. Win rate jumped from 0.42 to 0.85.",
"next_steps":["Try same constraints on Sonnet","Test without structure template"],
"figures":["figures/exp003_convergence.pdf"]
}
```
**Why a journal, not just git?** Git tracks file changes. The journal tracks the reasoning: why you tried X, what you learned, and what that implies for the next experiment. When writing the paper, this tree is invaluable for the Methods section ("we observed X, which motivated Y") and for honest failure reporting.
**Selecting the best path**: When the journal shows a branching tree (exp_001 → exp_002a, exp_002b, exp_003), identify the path that best supports the paper's claims. Document dead-end branches in the appendix as ablations or negative results.
**Snapshot code per experiment**: Copy the experiment script after each run:
- **Error bars**: Standard deviation or standard error, specify which
- **Confidence intervals**: 95% CI for key results
- **Pairwise tests**: McNemar's test for comparing two methods
- **Effect sizes**: Cohen's d or h for practical significance
See [references/experiment-patterns.md](references/experiment-patterns.md) for complete implementations of McNemar's test, bootstrapped CIs, and Cohen's h.
### Step 4.3: Identify the Story
After analysis, explicitly answer:
1.**What is the main finding?** State it in one sentence.
2.**What surprised you?** Unexpected results often make the best papers.
3.**What failed?** Failed experiments can be the most informative. Honest reporting of failures strengthens the paper.
4.**What follow-up experiments are needed?** Results often raise new questions.
#### Handling Negative or Null Results
When your hypothesis was wrong or results are inconclusive, you have three options:
| Situation | Action | Venue Fit |
|-----------|--------|-----------|
| Hypothesis wrong but **why** is informative | Frame paper around the analysis of why | NeurIPS, ICML (if analysis is rigorous) |
| Clean negative result on popular claim | Write it up — the field needs to know | NeurIPS Datasets & Benchmarks, TMLR, workshops |
| Results inconclusive, no clear story | Pivot — run different experiments or reframe | Don't force a paper that isn't there |
**How to write a negative results paper:**
- Lead with what the community believes and why it matters to test it
- Describe your rigorous methodology (must be airtight — reviewers will scrutinize harder)
- Present the null result clearly with statistical evidence
- Analyze **why** the expected result didn't materialize
- Discuss implications for the field
**Venues that explicitly welcome negative results**: NeurIPS (Datasets & Benchmarks track), TMLR, ML Reproducibility Challenge, workshops at major conferences. Some workshops specifically call for negative results.
### Step 4.4: Create Figures and Tables
**Figures**:
- Use vector graphics (PDF) for all plots: `plt.savefig('fig.pdf')`
- Colorblind-safe palettes (Okabe-Ito or Paul Tol)
- Self-contained captions — reader should understand without main text
- No title inside figure — the caption serves this function
| Results inconclusive, need more data | Back to Phase 2 (design) |
| Unexpected finding suggests new direction | Back to Phase 2 (design) |
| Missing one ablation reviewers will ask for | Run it, then Phase 5 |
| All experiments done but some failed | Note failures, move to Phase 5 |
### Step 4.6: Write the Experiment Log (Bridge to Writeup)
Before moving to paper writing, create a structured experiment log that bridges results to prose. This is the single most important connective tissue between experiments and the writeup — without it, the writing agent has to re-derive the story from raw result files.
**Create `experiment_log.md`** with the following structure:
```markdown
# Experiment Log
## Contribution (one sentence)
[The paper's main claim]
## Experiments Run
### Experiment 1: [Name]
- **Claim tested**: [Which paper claim this supports]
- **Setup**: [Model, dataset, config, number of runs]
| figures/main_comparison.pdf | Bar chart comparing all methods on benchmark X | Results, Figure 2 |
| figures/ablation.pdf | Ablation removing components A, B, C | Results, Figure 3 |
...
## Failed Experiments (document for honesty)
- [What was tried, why it failed, what it tells us]
## Open Questions
- [Anything the results raised that the paper should address]
```
**Why this matters**: When drafting, the agent (or a delegated sub-agent) can load `experiment_log.md` alongside the LaTeX template and produce a first draft grounded in actual results. Without this bridge, the writing agent must parse raw JSON/CSV files and infer the story — a common source of hallucinated or misreported numbers.
**Git discipline**: Commit this log alongside the results it describes.
---
## Iterative Refinement: Strategy Selection
Any output in this pipeline — paper drafts, experiment scripts, analysis — can be iteratively refined. The autoreason research provides empirical evidence for when each refinement strategy works and when it fails. Use this section to choose the right approach.
### Quick Decision Table
| Your Situation | Strategy | Why |
|---------------|----------|-----|
| Mid-tier model + constrained task | **Autoreason** | Sweet spot. Generation-evaluation gap is widest. Baselines actively destroy weak model outputs. |
| Mid-tier model + open task | **Autoreason** with scope constraints added | Add fixed facts, structure, or deliverable to bound the improvement space. |
| Frontier model + constrained task | **Autoreason** | Wins 2/3 constrained tasks even at frontier. |
| Frontier model + unconstrained task | **Critique-and-revise** or **single pass** | Autoreason comes last. Model self-evaluates well enough. |
| Concrete technical task (system design) | **Critique-and-revise** | Direct find-and-fix loop is more efficient. |
| Template-filling task (one correct structure) | **Single pass** or **conservative** | Minimal decision space. Iteration adds no value. |
| Code with test cases | **Autoreason (code variant)** | Structured analysis of *why* it failed before fixing. Recovery rate 62% vs 43%. |
| Very weak model (Llama 8B class) | **Single pass** | Model too weak for diverse candidates. Invest in generation quality. |
### The Generation-Evaluation Gap
**Core insight**: Autoreason's value depends on the gap between a model's generation capability and its self-evaluation capability.
```
Model Tier │ Generation │ Self-Eval │ Gap │ Autoreason Value
Weak (Llama 8B) │ Poor │ Poor │ Small │ None — can't generate diverse candidates
Mid (Haiku 3.5) │ Decent │ Poor │ LARGE │ MAXIMUM — 42/42 perfect Borda
Mid (Gemini Flash)│ Decent │ Moderate │ Large │ High — wins 2/3
Strong (Sonnet 4) │ Good │ Decent │ Medium │ Moderate — wins 3/5
Frontier (S4.6) │ Excellent │ Good │ Small │ Only with constraints
```
This gap is structural, not temporary. As costs drop, today's frontier becomes tomorrow's mid-tier. The sweet spot moves but never disappears.
### Autoreason Loop (Summary)
Each pass produces three candidates from fresh, isolated agents:
1.**Critic** → finds problems in incumbent A (no fixes)
2.**Author B** → revises A based on critique
3.**Synthesizer** → merges A and B (randomized labels)
4.**Judge Panel** → 3 blind CoT judges rank A, B, AB via Borda count
5.**Convergence** → A wins k=2 consecutive passes → done
**Key parameters:**
- k=2 convergence (k=1 premature, k=3 too expensive, no quality gain)
- CoT judges always (3x faster convergence)
- Temperature 0.8 authors, 0.3 judges
- Conservative tiebreak: incumbent wins ties
- Every role is a fresh agent with no shared context
### Applying to Paper Drafts
When refining the paper itself through autoreason:
- **Provide ground truth to the critic**: actual experimental data, result JSONs, statistical outputs. Without this, models hallucinate fabricated ablation studies and fake confidence intervals.
- **Use 3 working judges minimum**: A broken judge parser doesn't add noise — it prevents equilibrium entirely.
- **Scope constrain the revision**: "Address these specific weaknesses" not "improve the paper."
### Failure Modes
| Failure | Detection | Fix |
|---------|-----------|-----|
| No convergence (A never wins) | A wins <15% over 20+ passes | Add scope constraints to the task |
| Synthesis drift | Word counts grow unboundedly | Constrain structure and deliverable |
| Degradation below single pass | Baselines score higher than iterated output | Switch to single pass; model may be too weak |
| Overfitting (code) | High public-test pass, low private-test pass | Use structured analysis, not just test feedback |
See [references/autoreason-methodology.md](references/autoreason-methodology.md) for complete prompts, Borda scoring details, model selection guide, scope constraint design patterns, and compute budget reference.
---
## Phase 5: Paper Drafting
**Goal**: Write a complete, publication-ready paper.
### Context Management for Large Projects
A paper project with 50+ experiment files, multiple result directories, and extensive literature notes can easily exceed the agent's context window. Manage this proactively:
**What to load into context per drafting task:**
| Drafting Task | Load Into Context | Do NOT Load |
| Writing Introduction | `experiment_log.md`, contribution statement, 5-10 most relevant paper abstracts | Raw result JSONs, full experiment scripts, all literature notes |
| Writing Methods | Experiment configs, pseudocode, architecture description | Raw logs, results from other experiments |
| Writing Results | `experiment_log.md`, result summary tables, figure list | Full analysis scripts, intermediate data |
| Writing Related Work | Organized citation notes (Step 1.4 output), .bib file | Experiment files, raw PDFs |
| Revision pass | Full paper draft, specific reviewer concerns | Everything else |
**Principles:**
- **`experiment_log.md` is the primary context bridge** — it summarizes everything needed for writing without loading raw data files (see Step 4.6)
- **Load one section's context at a time** when delegating. A sub-agent drafting Methods doesn't need the literature review notes.
- **Summarize, don't include raw files.** For a 200-line result JSON, load a 10-line summary table. For a 50-page related paper, load the 5-sentence abstract + your 2-line note about its relevance.
- **For very large projects**: Create a `context/` directory with pre-compressed summaries:
figure_inventory.md # List of figures with descriptions
```
### The Narrative Principle
**The single most critical insight**: Your paper is not a collection of experiments — it's a story with one clear contribution supported by evidence.
Every successful ML paper centers on what Neel Nanda calls "the narrative": a short, rigorous, evidence-based technical story with a takeaway readers care about.
**Three Pillars (must be crystal clear by end of introduction):**
| Pillar | Description | Test |
|--------|-------------|------|
| **The What** | 1-3 specific novel claims | Can you state them in one sentence? |
| **The Why** | Rigorous empirical evidence | Do experiments distinguish your hypothesis from alternatives? |
| **The So What** | Why readers should care | Does this connect to a recognized community problem? |
**If you cannot state your contribution in one sentence, you don't yet have a paper.**
### The Sources Behind This Guidance
This skill synthesizes writing philosophy from researchers who have published extensively at top venues. The writing philosophy layer was originally compiled by [Orchestra Research](https://github.com/orchestra-research) as the `ml-paper-writing` skill.
| Source | Key Contribution | Link |
|--------|-----------------|------|
| **Neel Nanda** (Google DeepMind) | The Narrative Principle, What/Why/So What framework | [How to Write ML Papers](https://www.alignmentforum.org/posts/eJGptPbbFPZGLpjsp/highly-opinionated-advice-on-how-to-write-ml-papers) |
| **Sebastian Farquhar** (DeepMind) | 5-sentence abstract formula | [How to Write ML Papers](https://sebastianfarquhar.com/on-research/2024/11/04/how_to_write_ml_papers/) |
| **Gopen & Swan** | 7 principles of reader expectations | [Science of Scientific Writing](https://cseweb.ucsd.edu/~swanson/papers/science-of-writing.pdf) |
| **Zachary Lipton** | Word choice, eliminating hedging | [Heuristics for Scientific Writing](https://www.approximatelycorrect.com/2018/01/29/heuristics-technical-scientific-writing-machine-learning-perspective/) |
**Why?** Most reviewers form judgments before reaching your methods. Readers encounter your paper as: title → abstract → introduction → figures → maybe the rest.
### Writing Workflow
```
Paper Writing Checklist:
- [ ] Step 1: Define the one-sentence contribution
- [ ] Step 2: Draft Figure 1 (core idea or most compelling result)
- [ ] Step 9: Draft limitations (REQUIRED by all venues)
- [ ] Step 10: Plan appendix (proofs, extra experiments, details)
- [ ] Step 11: Complete paper checklist
- [ ] Step 12: Final review
```
### Two-Pass Refinement Pattern
When drafting with an AI agent, use a **two-pass** approach (proven effective in SakanaAI's AI-Scientist pipeline):
**Pass 1 — Write + immediate refine per section:**
For each section, write a complete draft, then immediately refine it in the same context. This catches local issues (clarity, flow, completeness) while the section is fresh.
**Pass 2 — Global refinement with full-paper context:**
After all sections are drafted, revisit each section with awareness of the complete paper. This catches cross-section issues: redundancy, inconsistent terminology, narrative flow, and gaps where one section promises something another doesn't deliver.
```
Second-pass refinement prompt (per section):
"Review the [SECTION] in the context of the complete paper.
- Does it fit with the rest of the paper? Are there redundancies with other sections?
- Is terminology consistent with Introduction and Methods?
- Can anything be cut without weakening the message?
- Does the narrative flow from the previous section and into the next?
Make minimal, targeted edits. Do not rewrite from scratch."
```
### LaTeX Error Checklist
Append this checklist to every refinement prompt. These are the most common errors when LLMs write LaTeX:
```
LaTeX Quality Checklist (verify after every edit):
- [ ] No unenclosed math symbols ($ signs balanced)
- [ ] Only reference figures/tables that exist (\ref matches \label)
- [ ] No fabricated citations (\cite matches entries in .bib)
- [ ] Every \begin{env} has matching \end{env} (especially figure, table, algorithm)
- [ ] No HTML contamination (</end{figure}> instead of \end{figure})
- [ ] No unescaped underscores outside math mode (use \_ in text)
- [ ] No duplicate \label definitions
- [ ] No duplicate section headers
- [ ] Numbers in text match actual experimental results
- [ ] All figures have captions and labels
- [ ] No overly long lines that cause overfull hbox warnings
```
### Step 5.0: Title
The title is the single most-read element of the paper. It determines whether anyone clicks through to the abstract.
**Good titles**:
- State the contribution or finding: "Autoreason: When Iterative LLM Refinement Works and Why It Fails"
- Highlight a surprising result: "Scaling Data-Constrained Language Models" (implies you can)
- Name the method + what it does: "DPO: Direct Preference Optimization of Language Models"
**Bad titles**:
- Too generic: "An Approach to Improving Language Model Outputs"
- Too long: anything over ~15 words
- Jargon-only: "Asymptotic Convergence of Iterative Stochastic Policy Refinement" (who is this for?)
**Rules**:
- Include your method name if you have one (for citability)
- Include 1-2 keywords reviewers will search for
- Avoid colons unless both halves carry meaning
- Test: would a reviewer know the domain and contribution from the title alone?
### Step 5.1: Abstract (5-Sentence Formula)
From Sebastian Farquhar (DeepMind):
```
1. What you achieved: "We introduce...", "We prove...", "We demonstrate..."
2. Why this is hard and important
3. How you do it (with specialist keywords for discoverability)
4. What evidence you have
5. Your most remarkable number/result
```
**Delete** generic openings like "Large language models have achieved remarkable success..."
### Step 5.2: Figure 1
Figure 1 is the second thing most readers look at (after abstract). Draft it before writing the introduction — it forces you to clarify the core idea.
| Figure 1 Type | When to Use | Example |
|---------------|-------------|---------|
| **Method diagram** | New architecture or pipeline | TikZ flowchart showing your system |
| **Results teaser** | One compelling result tells the whole story | Bar chart: "Ours vs baselines" with clear gap |
| **Problem illustration** | The problem is unintuitive | Before/after showing failure mode you fix |
**Rules**: Figure 1 must be understandable without reading any text. The caption alone should communicate the core idea. Use color purposefully — don't just decorate.
### Step 5.3: Introduction (1-1.5 pages max)
Must include:
- Clear problem statement
- Brief approach overview
- 2-4 bullet contribution list (max 1-2 lines each in two-column format)
- Methods should start by page 2-3
### Step 5.4: Methods
Enable reimplementation:
- Conceptual outline or pseudocode
- All hyperparameters listed
- Architectural details sufficient for reproduction
- Present final design decisions; ablations go in experiments
### Step 5.5: Experiments & Results
For each experiment, explicitly state:
- **What claim it supports**
- How it connects to main contribution
- What to observe: "the blue line shows X, which demonstrates Y"
Requirements:
- Error bars with methodology (std dev vs std error)
- Hyperparameter search ranges
- Compute infrastructure (GPU type, total hours)
- Seed-setting methods
### Step 5.6: Related Work
Organize methodologically, not paper-by-paper. Cite generously — reviewers likely authored relevant papers.
### Step 5.7: Limitations (REQUIRED)
All major conferences require this. Honesty helps:
- Reviewers are instructed not to penalize honest limitation acknowledgment
- Pre-empt criticisms by identifying weaknesses first
- The main paper must be self-contained — reviewers are not required to read appendices
- Never put critical evidence only in the appendix
- Cross-reference: "Full results in Table 5 (Appendix B)" not just "see appendix"
- Use `\appendix` command, then `\section{A: Proofs}` etc.
### Page Budget Management
When over the page limit:
| Cut Strategy | Saves | Risk |
|-------------|-------|------|
| Move proofs to appendix | 0.5-2 pages | Low — standard practice |
| Condense related work | 0.5-1 page | Medium — may miss key citations |
| Combine tables with subfigures | 0.25-0.5 page | Low — often improves readability |
| Use `\vspace{-Xpt}` sparingly | 0.1-0.3 page | Low if subtle, high if obvious |
| Remove qualitative examples | 0.5-1 page | Medium — reviewers like examples |
| Reduce figure sizes | 0.25-0.5 page | High — figures must remain readable |
**Do NOT**: reduce font size, change margins, remove required sections (limitations, broader impact), or use `\small`/`\footnotesize` for main text.
### Step 5.10: Ethics & Broader Impact Statement
Most venues now require or strongly encourage an ethics/broader impact statement. This is not boilerplate — reviewers read it and can flag ethics concerns that trigger desk rejection.
**What to include:**
| Component | Content | Required By |
|-----------|---------|-------------|
| **Positive societal impact** | How your work benefits society | NeurIPS, ICML |
print(f"Energy: {energy_kwh:.0f} kWh, Carbon: {carbon_kg:.0f} kg CO2eq")
```
### Step 5.11: Datasheets & Model Cards (If Applicable)
If your paper introduces a **new dataset** or **releases a model**, include structured documentation. Reviewers increasingly expect this, and NeurIPS Datasets & Benchmarks track requires it.
**Datasheets for Datasets** (Gebru et al., 2021) — include in appendix:
```
Dataset Documentation (Appendix):
- Motivation: Why was this dataset created? What task does it support?
- Composition: What are the instances? How many? What data types?
- Collection: How was data collected? What was the source?
- Preprocessing: What cleaning/filtering was applied?
- Distribution: How is the dataset distributed? Under what license?
- Maintenance: Who maintains it? How to report issues?
- Ethical considerations: Contains personal data? Consent obtained?
Potential for harm? Known biases?
```
**Model Cards** (Mitchell et al., 2019) — include in appendix for model releases:
```
Model Card (Appendix):
- Model details: Architecture, training data, training procedure
- Intended use: Primary use cases, out-of-scope uses
- Metrics: Evaluation metrics and results on benchmarks
- Ethical considerations: Known biases, fairness evaluations
- Limitations: Known failure modes, domains where model underperforms
Work through systematically: title/authors → abstract → introduction → methods → experiments → related work → conclusion → references → appendix. Compile after each section.
- Include direction symbols ($\uparrow$ higher better, $\downarrow$ lower better)
- Right-align numerical columns
- Consistent decimal precision
**Figures**:
- **Vector graphics** (PDF, EPS) for all plots and diagrams — `plt.savefig('fig.pdf')`
- **Raster** (PNG 600 DPI) only for photographs
- **Colorblind-safe palettes** (Okabe-Ito or Paul Tol)
- Verify **grayscale readability** (8% of men have color vision deficiency)
- **No title inside figure** — the caption serves this function
- **Self-contained captions** — reader should understand without main text
### Conference Resubmission
For converting between venues, see Phase 7 (Submission Preparation) — it covers the full conversion workflow, page-change table, and post-rejection guidance.
### Professional LaTeX Preamble
Add these packages to any paper for professional quality. They are compatible with all major conference style files:
```latex
% --- Professional Packages (add after conference style file) ---
# Available styles: 'science', 'ieee', 'nature', 'science+ieee'
# Add 'no-latex' if LaTeX is not installed on the machine generating plots
```
**Standard figure sizes** (two-column format):
- Single column: `figsize=(3.5, 2.5)` — fits in one column
- Double column: `figsize=(7.0, 3.0)` — spans both columns
- Square: `figsize=(3.5, 3.5)` — for heatmaps, confusion matrices
---
## Phase 6: Self-Review & Revision
**Goal**: Simulate the review process before submission. Catch weaknesses early.
### Step 6.1: Simulate Reviews (Ensemble Pattern)
Generate reviews from multiple perspectives. The key insight from automated research pipelines (notably SakanaAI's AI-Scientist): **ensemble reviewing with a meta-reviewer produces far more calibrated feedback than a single review pass.**
**Step 1: Generate N independent reviews** (N=3-5)
Use different models or temperature settings. Each reviewer sees only the paper, not other reviews. **Default to negative bias** — LLMs have well-documented positivity bias in evaluation.
```
You are an expert reviewer for [VENUE]. You are critical and thorough.
If a paper has weaknesses or you are unsure about a claim, flag it clearly
and reflect that in your scores. Do not give the benefit of the doubt.
Review this paper according to the official reviewer guidelines. Evaluate:
1. Soundness (are claims well-supported? are baselines fair and strong?)
2. Clarity (is the paper well-written? could an expert reproduce it?)
3. Significance (does this matter to the community?)
4. Originality (new insights, not just incremental combination?)
Each reviewer can refine their review after seeing the meta-review. Use an early termination sentinel: if the reviewer responds "I am done" (no changes), stop iterating.
**Model selection for reviewing**: Reviewing is best done with the strongest available model, even if you wrote the paper with a cheaper one. The reviewer model should be chosen independently from the writing model.
**Few-shot calibration**: If available, include 1-2 real published reviews from the target venue as examples. This dramatically improves score calibration. See [references/reviewer-guidelines.md](references/reviewer-guidelines.md) for example reviews.
### Step 6.1b: Visual Review Pass (VLM)
Text-only review misses an entire class of problems: figure quality, layout issues, visual consistency. If you have access to a vision-capable model, run a separate **visual review** on the compiled PDF:
```
You are reviewing the visual presentation of this research paper PDF.
Check for:
1. Figure quality: Are plots readable? Labels legible? Colors distinguishable?
2. Figure-caption alignment: Does each caption accurately describe its figure?
3. Layout issues: Orphaned section headers, awkward page breaks, figures far from their references
4. Table formatting: Aligned columns, consistent decimal precision, bold for best results
5. Visual consistency: Same color scheme across all figures, consistent font sizes
6. Grayscale readability: Would the figures be understandable if printed in B&W?
For each issue, specify the page number and exact location.
```
This catches problems that text-based review cannot: a plot with illegible axis labels, a figure placed 3 pages from its first reference, inconsistent color palettes between Figure 2 and Figure 5, or a table that's clearly wider than the column width.
### Step 6.1c: Claim Verification Pass
After simulated reviews, run a separate verification pass. This catches factual errors that reviewers might miss:
```
Claim Verification Protocol:
1. Extract every factual claim from the paper (numbers, comparisons, trends)
2. For each claim, trace it to the specific experiment/result that supports it
3. Verify the number in the paper matches the actual result file
4. Flag any claim without a traceable source as [VERIFY]
```
For agent-based workflows: delegate verification to a **fresh sub-agent** that receives only the paper text and the raw result files. The fresh context prevents confirmation bias — the verifier doesn't "remember" what the results were supposed to be.
### Step 6.2: Prioritize Feedback
After collecting reviews, categorize:
| Priority | Action |
|----------|--------|
| **Critical** (technical flaw, missing baseline) | Must fix. May require new experiments → back to Phase 2 |
| **High** (clarity issue, missing ablation) | Should fix in this revision |
| **Medium** (minor writing issues, extra experiments) | Fix if time allows |
| **Low** (style preferences, tangential suggestions) | Note for future work |
### Step 6.3: Revision Cycle
For each critical/high issue:
1. Identify the specific section(s) affected
2. Draft the fix
3. Verify the fix doesn't break other claims
4. Update the paper
5. Re-check against the reviewer's concern
### Step 6.4: Rebuttal Writing
When responding to actual reviews (post-submission), rebuttals are a distinct skill from revision:
**Format**: Point-by-point. For each reviewer concern:
```
> R1-W1: "The paper lacks comparison with Method X."
We thank the reviewer for this suggestion. We have added a comparison with
Method X in Table 3 (revised). Our method outperforms X by 3.2pp on [metric]
(p<0.05). We note that X requires 2x our compute budget.
```
**Rules**:
- Address every concern — reviewers notice if you skip one
- Lead with the strongest responses
- Be concise and direct — reviewers read dozens of rebuttals
- Include new results if you ran experiments during the rebuttal period
- Never be defensive or dismissive, even of weak criticisms
- Use `latexdiff` to generate a marked-up PDF showing changes (see Professional LaTeX Tooling section)
- Thank reviewers for specific, actionable feedback (not generic praise)
**What NOT to do**: "We respectfully disagree" without evidence. "This is out of scope" without explanation. Ignoring a weakness by only responding to strengths.
### Step 6.5: Paper Evolution Tracking
Save snapshots at key milestones:
```
paper/
paper.tex # Current working version
paper_v1_first_draft.tex # First complete draft
paper_v2_post_review.tex # After simulated review
paper_v3_pre_submission.tex # Final before submission
paper_v4_camera_ready.tex # Post-acceptance final
```
---
## Phase 7: Submission Preparation
**Goal**: Final checks, formatting, and submission.
### Step 7.1: Conference Checklist
Every venue has mandatory checklists. Complete them carefully — incomplete checklists can result in desk rejection.
See [references/checklists.md](references/checklists.md) for:
- NeurIPS 16-item paper checklist
- ICML broader impact + reproducibility
- ICLR LLM disclosure policy
- ACL mandatory limitations section
- Universal pre-submission checklist
### Step 7.2: Anonymization Checklist
Double-blind review means reviewers cannot know who wrote the paper. Check ALL of these:
```
Anonymization Checklist:
- [ ] No author names or affiliations anywhere in the PDF
- [ ] No acknowledgments section (add after acceptance)
- [ ] Self-citations written in third person: "Smith et al. [1] showed..." not "We previously showed [1]..."
- [ ] No GitHub/GitLab URLs pointing to your personal repos
- [ ] Use Anonymous GitHub (https://anonymous.4open.science/) for code links
- [ ] No institutional logos or identifiers in figures
- [ ] No file metadata containing author names (check PDF properties)
- [ ] No "our previous work" or "in our earlier paper" phrasing
- [ ] Supplementary materials don't contain identifying information
```
**Common mistakes**: Git commit messages visible in supplementary code, watermarked figures from institutional tools, acknowledgments left in from a previous draft, arXiv preprint posted before anonymity period.
### Step 7.3: Formatting Verification
```
Pre-Submission Format Check:
- [ ] Page limit respected (excluding references and appendix)
- [ ] All figures are vector (PDF) or high-res raster (600 DPI PNG)
- [ ] All figures readable in grayscale
- [ ] All tables use booktabs
- [ ] References compile correctly (no "?" in citations)
| Any → COLM | varies → 9 | Reframe for language model focus |
When cutting pages: move proofs to appendix, condense related work, combine tables, use subfigures.
When expanding: add ablations, expand limitations, include additional baselines, add qualitative examples.
**After rejection**: Address reviewer concerns in the new version, but don't include a "changes" section or reference the previous submission (blind review).
Posting to arXiv is standard practice in ML but has important timing and anonymity considerations.
**Timing decision tree:**
| Situation | Recommendation |
|-----------|---------------|
| Submitting to double-blind venue (NeurIPS, ICML, ACL) | Post to arXiv **after** submission deadline, not before. Posting before can technically violate anonymity policies, though enforcement varies. |
| Submitting to ICLR | ICLR explicitly allows arXiv posting before submission. But don't put author names in the submission itself. |
| Paper already on arXiv, submitting to new venue | Acceptable at most venues. Do NOT update arXiv version during review with changes that reference reviews. |
| Workshop paper | arXiv is fine at any time — workshops are typically not double-blind. |
| Want to establish priority | Post immediately if scooping is a concern — but accept the anonymity tradeoff. |
**arXiv category selection** (ML/AI papers):
| Category | Code | Best For |
|----------|------|----------|
| Machine Learning | `cs.LG` | General ML methods |
| Computation and Language | `cs.CL` | NLP, language models |
| **Short paper** | 4 | Focused contribution: one clear point with evidence |
| **Findings** | 8 | Solid work that narrowly missed main conference |
**Short paper strategy**: Pick ONE claim and support it thoroughly. Don't try to compress a long paper into 4 pages — write a different, more focused paper.
---
## Paper Types Beyond Empirical ML
The main pipeline above targets empirical ML papers. Other paper types require different structures and evidence standards. See [references/paper-types.md](references/paper-types.md) for detailed guidance on each type.
### Theory Papers
**Structure**: Introduction → Preliminaries (definitions, notation) → Main Results (theorems) → Proof Sketches → Discussion → Full Proofs (appendix)
**Key differences from empirical papers:**
- Contribution is a theorem, bound, or impossibility result — not experimental numbers
- Methods section replaced by "Preliminaries" and "Main Results"
- Proofs are the evidence, not experiments (though empirical validation of theory is welcome)
- Proof sketches in main text, full proofs in appendix is standard practice
- Experimental section is optional but strengthens the paper if it validates theoretical predictions
**Proof writing principles:**
- State theorems formally with all assumptions explicit
- Provide intuition before formal proof ("The key insight is...")
- Proof sketches should convey the main idea in 0.5-1 page
- Use `\begin{proof}...\end{proof}` environments
- Number assumptions and reference them in theorems: "Under Assumptions 1-3, ..."
| **data-science** | Phase 4 (Analysis): Jupyter live kernel for interactive analysis and visualization | `skill_view("data-science")` |
**This skill supersedes `ml-paper-writing`** — it contains all of ml-paper-writing's content plus the full experiment/analysis pipeline and autoreason methodology.
| **`execute_code`** | Run Python for citation verification, statistical analysis, data aggregation. Has tool access via RPC. |
| **`read_file`** / **`write_file`** / **`patch`** | Paper editing, experiment scripts, result files. Use `patch` for targeted edits to large .tex files. |
| **`web_search`** | Literature discovery: `web_search("transformer attention mechanism 2024")` |
| **`web_extract`** | Fetch paper content, verify citations: `web_extract("https://arxiv.org/abs/2303.17651")` |
| **`delegate_task`** | **Parallel section drafting** — spawn isolated subagents for each section. Also for concurrent citation verification. |
| **`todo`** | Primary state tracker across sessions. Update after every phase transition. |
5. Report: table of results, key finding, next step
6. If nothing changed: respond with [SILENT]"
})
```
**[SILENT] protocol**: When nothing has changed since the last check, respond with exactly `[SILENT]`. This suppresses notification delivery to the user. Only report when there are genuine changes worth knowing about.
**Deadline tracking**:
```
cronjob("create", {
"schedule": "0 9 * * *", # Daily at 9am
"prompt": "NeurIPS 2025 deadline: May 22. Today is {date}.
Days remaining: {compute}.
Check todo list — are we on track?
If <7 days: warn user about remaining tasks."
})
```
### Communication Patterns
**When to notify the user** (via `send_message` or direct response):
- Experiment batch completed (with results table)
- Unexpected finding or failure requiring decision
- Draft section ready for review
- Deadline approaching with incomplete tasks
**When NOT to notify:**
- Experiment still running, no new results → `[SILENT]`
- Routine monitoring with no changes → `[SILENT]`
- Intermediate steps that don't need attention
**Report format** — always include structured data:
```
## Experiment: <name>
Status: Complete / Running / Failed
| Task | Method A | Method B | Method C |
|------|---------|---------|---------|
| Task 1 | 85.2 | 82.1 | **89.4** |
Key finding: <one sentence>
Next step: <what happens next>
```
### Decision Points Requiring Human Input
Use `clarify` for targeted questions when genuinely blocked:
| Decision | When to Ask |
|----------|-------------|
| Target venue | Before starting paper (affects page limits, framing) |
| Contribution framing | When multiple valid framings exist |
| Experiment priority | When TODO list has more experiments than time allows |
| Submission readiness | Before final submission |
**Do NOT ask about** (be proactive, make a choice, flag it):
- Word choice, section ordering
- Which specific results to highlight
- Citation completeness (draft with what you find, note gaps)
---
## Reviewer Evaluation Criteria
Understanding what reviewers look for helps focus effort:
- 1: Strong Reject — known results or ethics issues
See [references/reviewer-guidelines.md](references/reviewer-guidelines.md) for detailed guidelines, common concerns, and rebuttal strategies.
---
## Common Issues and Solutions
| Issue | Solution |
|-------|----------|
| Abstract too generic | Delete first sentence if it could prepend any ML paper. Start with your specific contribution. |
| Introduction exceeds 1.5 pages | Split background into Related Work. Front-load contribution bullets. |
| Experiments lack explicit claims | Add: "This experiment tests whether [specific claim]..." before each one. |
| Reviewers find paper hard to follow | Add signposting, use consistent terminology, make figure captions self-contained. |
| Missing statistical significance | Add error bars, number of runs, statistical tests, confidence intervals. |
| Scope creep in experiments | Every experiment must map to a specific claim. Cut experiments that don't. |
| Paper rejected, need to resubmit | See Conference Resubmission in Phase 7. Address reviewer concerns without referencing reviews. |
| Missing broader impact statement | See Step 5.10. Most venues require it. "No negative impacts" is almost never credible. |
| Human eval criticized as weak | See Step 2.5 and [references/human-evaluation.md](references/human-evaluation.md). Report agreement metrics, annotator details, compensation. |
| Reviewers question reproducibility | Release code (Step 7.9), document all hyperparameters, include seeds and compute details. |
| Theory paper lacks intuition | Add proof sketches with plain-language explanations before formal proofs. See [references/paper-types.md](references/paper-types.md). |
| Results are negative/null | See Phase 4.3 on handling negative results. Consider workshops, TMLR, or reframing as analysis. |
