Jacobian Scopes: A Unified Geometric Framework for Token-Level LLM Attribution
cs.CL, cs.AI
Submitted: 2026-01-23
Updated: 2026-10-27
Comments: 35 pages, 17 figures
Journal ref: Published as a conference paper at Neurips 2026
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- What Can Transformers Learn In-Context? A Case Study of Simple Function Classes
- The Llama 3 Herd of Models
- Bias in Large Language Models: Origin, Evaluation, and Mitigation
- How to use and interpret activation patching
- Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2
- In-context Learning and Induction Heads
- Qwen2.5 Technical Report
- Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps
- SmoothGrad: removing noise by adding noise
- Gemma 3 Technical Report
- A Stitch in Time Saves Nine: Detecting and Mitigating Hallucinations of LLMs by Validating Low-Confidence Generation
- Context parroting: A simple but tough-to-beat baseline for foundation models in scientific machine learning
Related papers
- ALARM: Audio-Language Alignment for Reasoning Models
- ThinkSum: Probabilistic reasoning over sets using large language models
- Coherence boosting: When your pretrained language model is not paying enough attention
- From Facts to Insights: A Persona-Driven Dual Memory Framework and Dataset for Role-Playing Agents
- Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization
- The Art of Asking: Multilingual Prompt Optimization for Synthetic Data