Showing 120 of 120on this page. Filters & sort apply to loaded results; URL updates for sharing.120 of 120 on this page
Understanding SAE Features with the Logit Lens — AI Alignment Forum
interpreting GPT: the logit lens — LessWrong
Protein language models through the logit lens | Liam Bai
interpreting GPT: the logit lens — AI Alignment Forum
Decoding an LLM’s Thoughts: Logit Lens in Just 25 Lines of Code | by ...
The logit lens can be deceptive if not used properly
Understanding SAE Features with the Logit Lens — LessWrong
Interpreting GPT-2 Brain using Logit Lens | by Chamod Kalupahana | Medium
2138 Beyond Logit Lens Context | PDF | Image Segmentation
Probing CODI's Latent Reasoning Chain with Logit Lens and Tuned Lens ...
(PDF) LogitLens4LLMs: Extending Logit Lens Analysis to Modern Large ...
[Proposal] Logit Lens Analysis for Gemma 2B (Ref #154) · Issue #1090 ...
Workbench Logit Lens Demo - YouTube
Logit Lens & ViT model: туториал / Хабр
Neural Mechanics Week 1: LLM Foundations and Logit Lens with David Bau ...
Recently re-implemented GPT Logit Lens (https://lnkd.in/erYfxjAD) It is ...
Gen_AI 补充内容 Logit Lens 和 Patchscopes-CSDN博客
Logit Obiektyw Do Telefonu Apexel 6W1 Makro Szerokokątny Lens Kit Wide ...
Blog post "interpreting GPT: the logit lens" demonstrates GPT-2's most ...
Understanding Logit Lens: Peering into Language Model Layers | by ...
Logit Lens: Interpreting Neural Logits
Figure 2 from A Logit Adjusting Transformer for Class Imbalance in ...
Figure 3 from Beyond Logit Lens: Contextual Embeddings for Robust ...
LLM Interpretability: Logit Lens. Hopefully you made it through the 35 ...
Paper page - Beyond Logit Lens: Contextual Embeddings for Robust ...
Logit Prisms: Decomposing Transformer Outputs for Mechanistic ...
The estimation results of the logit model (Part 1) | Download ...
Beyond Logit Lens: Contextual Embeddings for Robust Hallucination ...
探索Transformer的内部机制：Tuned Lens方法详解-CSDN博客
Gears-Level Mental Models of Transformer Interpretability — AI ...
Structure and Interpretation of Deep Networks
RevLLM
Logit-Lens-in-LMMs-Research-Work/LogitLens_for_Qwen2_VL_T_bank_.ipynb ...
Unlocking Visual Insights: Applying the ‘Logit Lens’ to Image Data with ...
Tutorials — nnsight
Week 1: Foundations - Neural Mechanics
Laying the Foundations for Vision and Multimodal Mechanistic ...
Python实现离散选择Logit模型(Logit算法)项目实战_python logit模型-CSDN博客
Future Lens: Anticipating Subsequent Tokens from a Single Hidden State
Logit-Lens/logit_lens.ipynb at main · SullivanCastro/Logit-Lens · GitHub
Really Strong Features Found in Residual Stream — LessWrong
Small Models Can Introspect, Too
GitHub - zhenyu-02/LogitLens4LLMs: A versatile toolkit for applying ...
We Found An Neuron in GPT-2 - Clement Neo
无需思维链，LLM如何“凭空”算出答案？探究末位Token的计算机制 - 知乎
通过删除LLM中的敏感信息保护隐私可能并不完美 - LLM Safety论文精读（六） - 知乎
Beyond the Black Box: Interpretability of LLMs in Finance | AI Research ...
手搓GPT系列之 - 如何理解logit函数-CSDN博客
Representation Tuning — LessWrong
Inside the mind of a superhuman Go model: How does Leela Zero read ...
NeurIPS 2025 | 医疗 AI LogitGaze-Med：重新定义医学影像注意力预测 - 知乎
