Research map / Large language models / Language, safety and interpretability
Hessian-Enhanced Token Attribution (HETA): Interpreting Autoregressive LLMs
Vishal Pramanik, Maisha Maliha, Nathaniel D. Bastian, Sumit Kumar Jha
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 |
|---|---|---|---|
| use HETA instead of ReAgent (next best) for attribution faithfulness | decoder-only LMs · LongRA Soft-NC, GPT-J 6B | Soft-NC: 1.68 -> 10.3 | Improved |
| use HETA instead of ReAgent on curated DSA alignment | decoder-only LMs · Curated DSA, GPT-J | DSA: 3.60 -> 4.80 | Improved |
| use HETA instead of baselines under decoding hyperparameter sweep | Llama-3.1 70B · Soft-NC sensitivity to decoding settings | max relative change (%): 2.3 -> 0.6 | Improved |
| remove semantic transition vector or replace with uniform weighting (ablation) | HETA · LongRA, TellMeWhy, WikiBio, curated (GPT-J 6B) | Soft-NC, Soft-NS, DSA | Improved |
| add Hessian-based sensitivity and KL information terms to transition (ablation) | HETA · LongRA, TellMeWhy, WikiBio, curated (GPT-J 6B) | Soft-NC, Soft-NS, DSA | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ContextCite — ContextCite: Attributing Model Generation to Context (NeurIPS 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Scalable Influence and Fact Tracing for Large Language Model Pretraining (ICLR 2025 · similarity 0.88)
- AttriBoT: A Bag of Tricks for Efficiently Approximating Leave-One-Out Context Attribution (ICLR 2025 · similarity 0.88)
- Flexible Language Modeling in Continuous Space with Transformer-based Autoregressive Flows (NeurIPS 2025 · similarity 0.87)
- Sketching the Readout of Large Language Models for Scalable Data Attribution and Valuation (NeurIPS 2026 · similarity 0.87)
- DATE-LM: Benchmarking Data Attribution Evaluation for Large Language Models (NeurIPS 2025 · similarity 0.87)
- Fast and Fluent Diffusion Language Models via Convolutional Decoding and Rejective Fine-tuning (NeurIPS 2025 · similarity 0.87)
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.
