Register and share your invite link to earn from video plays and referrals.

Grigory Sapunov
@che_shr_cat
PhD in AI | GDE in AI | CTO Intento | Author "Deep Learning with JAX" ๐Ÿ“ ML insights: ๐Ÿค– AI reviews:
1.8K Following    4.6K Followers
1/ From 7B to 30B parameters, zeroing fewer than 0.05% of frozen weights with GRPO matches full fine-tuning on math, and on Qwen2.5-7B it beats the base model even at pass@128 (80.08% vs 74.01%) without writing a single new weight. ๐Ÿงต
Show more
2/ Specification gaming is not just genetic algorithm folklore anymore. In "AI Finds A Way", Aaron Dharna, Cong Lu, Ryan Sullivan, Joel Lehman, @vkrakovna, and @jeffclune catalog 26 case studies proving this behavior is universal across deep RL, LLMs, and AI for science.
Show more
1/ SSL models like DINOv2 are engineering marvels, but they rely on fragile hacks like EMA, stop-gradients, and custom centering to avoid representation collapse. What if we could replace all these heuristics with pure, mathematically sound optimal transport? ๐Ÿงต
Show more