Blog
Welcome to my blog! Here I share insights from my daily readings of AI research papers and write about interesting open-source projects I’m working on. I hope these posts are helpful to the community and spark interesting discussions.
AI Daily Paper Summary
I read AI and ML research papers daily and post concise summaries here. Each entry covers the key contributions, methods, and takeaways from recent papers across topics like deep learning, reinforcement learning, NLP, computer vision, and more.
Latest — October 18, 2026
- LeanPRM — Uses Lean-grounded supervision to train more accurate natural-language proof verifiers. arxiv.org/abs/2610.17697
- Copy-Paste LMs — Adds an explicit copying head that improves Transformer training and contextual retrieval. arxiv.org/abs/2610.19247
- Verifier Hardening — Iteratively attacks and repairs reward verifiers to reduce agent reward hacking. arxiv.org/abs/2610.15924
- Recursive Environment Scaling — Creates harder long-horizon agent tasks while preserving their original objectives. arxiv.org/abs/2610.16397
- Unembedding Dynamics — Reveals that models can encode capabilities well before their output layers expose them. arxiv.org/abs/2610.21025
- PathFM — Generates coherent multivariate futures while keeping trajectory sampling fast. arxiv.org/abs/2610.22358
- MoE Personalization — Identifies and causally controls experts responsible for personalized language-model behavior. arxiv.org/abs/2610.22127
- H-DDCM — Makes generative image and video compression practical with efficient Hadamard codebooks. arxiv.org/abs/2610.18125
- Speech Event Detector — Adds efficient timestamped non-verbal events and background sounds to speech transcripts. arxiv.org/abs/2610.14232
- CoWorld Reasoner — Turns video world models into selectable visual-imagination tools for multimodal reasoning. arxiv.org/abs/2610.21910
Interesting Projects
In this section, I share outcomes, results, and lessons learned from open-source projects I’m working on. The goal is to give back to the community and document the journey of building and experimenting with different ideas.