Writing Philosophy for Academic Papers

"A paper is a short, rigorous, evidence-based technical story with a takeaway readers care about." — Neel Nanda

Table of Contents

• The Narrative Principle
• 5-Sentence Abstract Formula
• Gopen-Swan 7 Principles
• Word Choice
• Precision Over Brevity
• Micro-Level Tips
• Section-by-Section Guide
• Time Allocation

The Narrative Principle

(Neel Nanda, Google DeepMind)

Your paper is NOT a collection of experiments — it's a story with ONE clear contribution.

Three Pillars (must be crystal clear by end of introduction):

Pillar	Description	Example
The What	1-3 specific novel claims within cohesive theme	"We prove that X achieves Y under condition Z"
The Why	Rigorous empirical evidence supporting claims	Strong baselines, experiments distinguishing hypotheses
The So What	Why readers should care	Connection to recognized community problems

If you cannot state your contribution in one sentence, you don't yet have a paper.

5-Sentence Abstract Formula

(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..."

Gopen-Swan 7 Principles

(Based on "The Science of Scientific Writing")

#	Principle	Rule	Example
1	Subject-verb proximity	Keep subject and verb close	❌ "The model, which was trained on..., achieves" → ✅ "The model achieves... after training on..."
2	Stress position	Place emphasis at sentence ends	❌ "Accuracy improves by 15% when using attention" → ✅ "When using attention, accuracy improves by 15%"
3	Topic position	Put context first, new info after	✅ "Given these constraints, we propose..."
4	Old before new	Familiar info → unfamiliar info	Link backward, then introduce new
5	One unit, one function	Each paragraph makes one point	Split multi-point paragraphs
6	Action in verb	Use verbs, not nominalizations	❌ "We performed an analysis" → ✅ "We analyzed"
7	Context before new	Set stage before presenting	Explain before showing equation

Word Choice

(Zachary Lipton)

Be Specific

• ❌ "performance" → ✅ "accuracy" or "latency" (say what you mean)
• ❌ "significant" → ✅ "statistically significant (p < 0.05)" or remove

Eliminate Hedging

Drop "may" and "can" unless genuinely uncertain.

• ❌ "This may improve performance" → ✅ "This improves accuracy by 3.2%"

Avoid Incremental Vocabulary

• ❌ "combine," "modify," "expand" → ✅ "develop," "propose," "introduce"

Delete Intensifiers

• ❌ "provides very tight approximation" → ✅ "provides tight approximation"

Precision Over Brevity

(Jacob Steinhardt, UC Berkeley)

• Consistent terminology: Different terms for same concept creates confusion. Pick one and stick with it.
• State assumptions formally: Before theorems, list all assumptions explicitly
• Intuition + rigor: Provide intuitive explanations alongside formal proofs
• Define before use: Every symbol and term defined before first use

Micro-Level Tips

(Ethan Perez, Anthropic)

• [ ] Minimize pronouns: ❌ "This shows..." → ✅ "This result shows..."
• [ ] Verbs early: Position verbs near sentence start
• [ ] Unfold apostrophes: ❌ "X's Y" → ✅ "The Y of X" (when awkward)
• [ ] Delete filler words: "actually," "a bit," "very," "really," "basically," "quite," "essentially"
• [ ] Avoid vague "this": Always attach a noun after "this"
• [ ] Active voice: ❌ "The method was applied" → ✅ "We applied the method"

Section-by-Section Guide

Section	Length	Key Requirements
Abstract	150-250 words	5-sentence formula; delete generic openings
Introduction	1-1.5 pages max	2-4 bullet contribution list; methods should start by page 2-3
Methods	As needed	Enable reimplementation; all hyperparameters listed
Experiments	Main body	State claim before each experiment; error bars mandatory
Related Work	1-1.5 pages	Organize methodologically; each cited work needs individual discussion; max 2 clustered citations
Limitations	0.5-1 page	Required by all major venues; pre-empt reviewer criticisms

Introduction Must Include:

• Clear problem statement
• 2-4 bullet contribution list (max 1-2 lines each)
• Brief approach overview
• Citation integration: Each cited work must be accompanied by a brief description of its contribution, method, or limitation — max 2 clustered citations without discussion (see STYLE_GUIDE.md Anti-Citation-Stacking Rules)

Experiments Must Include:

• What claim each experiment supports
• Error bars with methodology (std dev vs std error)
• Hyperparameter search ranges
• Compute infrastructure (GPU type, total hours)

Time Allocation

(Neel Nanda)

Spend approximately equal time on each of:

1. The abstract
2. The introduction
3. The figures
4. Everything else combined

Why? Most reviewers form judgments before reaching your methods.

Reader encounter order: title → abstract → introduction → figures → maybe the rest.

Sources

Source	Key Contribution	Link
Neel Nanda (Google DeepMind)	The Narrative Principle	How to Write ML Papers
Sebastian Farquhar (DeepMind)	5-sentence abstract formula	How to Write ML Papers
Gopen & Swan	7 principles of reader expectations	Science of Scientific Writing
Zachary Lipton	Word choice, eliminating hedging	Heuristics for Scientific Writing
Jacob Steinhardt (UC Berkeley)	Precision, consistent terminology	Writing Tips
Ethan Perez (Anthropic)	Micro-level clarity tips	Easy Paper Writing Tips
Andrej Karpathy	Single contribution focus	Various lectures