LensLapse: In-Browser Logit Lens for Interactive Exploration of Language Model Training Dynamics

Tatsuki Okada, Yusuke Sakai and Taro Watanabe presented LensLapse, an MIT-licensed open-source web application that lets users scrub a slider across public checkpoints from five model families and three multilingual checkpoint suites to watch per-layer, per-position logit lens predictions shift from frequency bias to factual knowledge. For browser-compatible models, arbitrary user prompts are probed live in the browser via local inference over per-checkpoint model pairs, with no server, API key or hosting cost, and prompts never leaving the user's device. A single pip install lets users launch the application locally to explore their own models and checkpoints; the paper appears in the AACL-IJCNLP 2026 System Demonstrations proceedings, pages 66–73, and the tool is hosted at iamtatsuki05.github.io/lenslapse.

Abstract The logit lens is a standard mechanistic interpretability tool for inspecting a Transformer’s predictions across layers. However, existing logit lens tools typically focus on individual checkpoints rather than interactive exploration of a model’s training history . In this study, we introduce LensLapse , an open-source web application for interactive exploration of a model’s training history through token-level interpretability. Users can scrub a slider across public checkpoints from five model families and three multilingual checkpoint suites and observe how per-layer, per-position logit lens predictions evolve from frequency bias to factual knowledge. For browser-compatible models, arbitrary user prompts are probed live in the browser via local inference over per-checkpoint model pairs. No server, no API key, no hosting cost, and prompts never leave the user’s device. Moreover, LensLapse supports local models. A single ‘pip install‘ enables users to launch the application and explore their own models and training checkpoints without any additional web setup. LensLapse is MIT-licensed and permanently hosted at iamtatsuki05.github.io/lenslapse. - GitHub: https://github.com/iamtatsuki05/lenslapse - PyPI: https://pypi.org/project/lenslapse/ - docs: https://iamtatsuki05.github.io/lenslapse/docs/ - YouTube: https://youtu.be/OBnvG9Ru3j8 - Anthology ID: - 2026.aacl-demo.9 - Volume: - Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics: System Demonstrations https://aclanthology.org/volumes/2026.aacl-demo/ - Month: - November - Year: - 2026 - Address: - Hengqin, China - Editors: - Ningyu Zhang https://aclanthology.org/people/ningyu-zhang/ , Derry Wijaya https://aclanthology.org/people/derry-tanti-wijaya/ - Venues: - AACL https://aclanthology.org/venues/aacl/ | IJCNLP https://aclanthology.org/venues/ijcnlp/ - SIG: - Publisher: - Association for Computational Linguistics - Note: - Pages: - 66–73 - Language: - URL: - https://aclanthology.org/2026.aacl-demo.9/ https://aclanthology.org/2026.aacl-demo.9/ - DOI: - Cite ACL : - Tatsuki Okada, Yusuke Sakai, and Taro Watanabe. 2026. LensLapse: In-Browser Logit Lens for Interactive Exploration of Language Model Training Dynamics https://aclanthology.org/2026.aacl-demo.9/ . In Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics: System Demonstrations , pages 66–73, Hengqin, China. Association for Computational Linguistics. - Cite Informal : - LensLapse: In-Browser Logit Lens for Interactive Exploration of Language Model Training Dynamics https://aclanthology.org/2026.aacl-demo.9/ Okada et al., AACL-IJCNLP 2026 - PDF: - https://aclanthology.org/2026.aacl-demo.9.pdf https://aclanthology.org/2026.aacl-demo.9.pdf