Research map / Large language models / Language, safety and interpretability
How does GPT-2 compute greater-than?: Interpreting mathematical abilities in a pre-trained language model
Michael Hanna, Ollie Liu, Alexandre Variengien
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| corrupt circuit nodes while the rest of the model sees normal input (ablation) | GPT-2 small · Year-span prediction | probability difference: 81.7% -> -36.6% | Hurt |
| remove linearly extractable YY information with LEACE from attention head output (ablation) | GPT-2 small · Year-span prediction | probability difference: 81.7% -> 64.7% | Hurt |
| ablate PCA dimensions of structured number representations (ablation) | GPT-2 small · Year-span prediction | probability difference | No clear effect |
| patch MLP 8 direct path vs indirect paths (ablation) | GPT-2 small · Year-span prediction | performance drop: 14% direct vs 39% indirect | Mixed |
| apply year-span circuit to started/ended and price-range prompts | year-span circuit · related greater-than prompts | loss recovery: 98.8% and 88.9% | Improved |
| add MLP 7 and two extra attention heads to the circuit for number-sequence prompt (ablation) | year-span circuit · 1599, 1607, 1633, 1679, 17YY, 17 | performance recovery: 67.8% -> 90.3% | Improved |
Closest papers by meaning
- Interpreting and Improving Large Language Models in Arithmetic Calculation (ICML 2024 · similarity 0.87)
- Does GPT Really Get It? A Hierarchical Scale to Quantify Human and AI’s Understanding of Algorithms (AAAI 2025 · similarity 0.87)
- Arithmetic Without Algorithms: Language Models Solve Math with a Bag of Heuristics (ICLR 2025 · similarity 0.87)
- Language Models Trained to do Arithmetic Predict Human Risky and Intertemporal Choice (ICLR 2025 · similarity 0.86)
- Understanding Emergent Abilities of Language Models from the Loss Perspective (NeurIPS 2024 · similarity 0.86)
- Rethinking Fine-Tuning when Scaling Test-Time Compute: Limiting Confidence Improves Mathematical Reasoning (NeurIPS 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
