Search papers, labs, and topics across Lattice.
2
3
6
2
The DeepSeek-V4.1-Flash model, a multimodal Mixture-of-Experts model with 552B backbone parameters and support for contexts of up to one million tokens, is introduced, substantially improving cost efficiency for agentic workloads and pushing the limits of KV cache compression.
Standard GRPO's uniform clipping boundary actively chokes exploration by penalizing rare breakthroughs on hard problems just as harshly as trivial rollouts on easy ones.
