https://apurvanakade.github.io/blog/tags/ml/ Recent content in ML on PostDoc Problems Hugo en Sat, 09 Sep 2023 22:37:41 +0000 https://apurvanakade.github.io/blog/maths--science/lean/2023-09-09-llm/ Sat, 09 Sep 2023 22:37:41 +0000 https://apurvanakade.github.io/blog/maths--science/lean/2023-09-09-llm/ <p>I came across two LLM based tactics for Lean</p> <ul> <li><a href="https://github.com/wellecks/llmstep">llmstep: [L]LM proofstep suggestions in Lean</a></li> <li><a href="https://github.com/lean-dojo/LeanInfer">LeanInfer: Neural Network Inference in Lean 4</a></li> </ul> <p>They both use [LeanDojo](2023-09-04 leandojo.qmd) somehow and I don&rsquo;t understand the difference between the two. Unfortunately, I do not know enough about LLM and AI to tell if this is actually good or just hot gas.</p> https://apurvanakade.github.io/blog/maths--science/lean/2023-09-04-leandojo/ Mon, 04 Sep 2023 16:22:19 +0000 https://apurvanakade.github.io/blog/maths--science/lean/2023-09-04-leandojo/ <p>This was posted a while back on the Lean Zulip:</p> <p><a href="https://leandojo.org/">LeanDojo</a></p> <p><a href="https://arxiv.org/abs/2306.15626">LeanDojo: Theorem Proving with Retrieval-Augmented Language Models </a></p> <p>According to their description, LeanDojo is an open-source Lean playground consisting of toolkits, data, models, and benchmarks. Plus they also provide an LLM-based prover and a chatGPT plugin. All of this looks amazing and something worth learning more about.</p>