Explore tweets tagged as #InferenceTimeCompute
@xlzhao_hku
Xueliang Zhao
2 months
⚙️ We integrate vLLM into MCTS, accelerating node expansion, rollout, and reward evaluation. Beyond reasoning accuracy, this creates a reusable acceleration framework for large-scale MCTS and TTS research. #MCTS #vLLM #InferenceTimeCompute
1
0
3
@lukeprog
Luke Muehlhauser
9 months
Cool new RFP from Schmidt Sciences; apply by Apr 30! https://t.co/mWznoZW7WJ
0
8
24
@asankhaya
Asankhaya Sharma
5 months
@chl260 8/8 🔧 OptiLLM continues to implement cutting-edge inference techniques: - MoA, AutoThink for reasoning - SPL for system prompt learning - Memory for unbounded context - And now TTD-DR for deep research! All open source: https://t.co/p7B40LSFOX #AI #LLM #InferenceTimeCompute
0
0
10
@emiyazono
Evan Miyazono
8 months
Opportunity from @schmidtsciences Up to $500K grants to 12-18 month projects improving safety in the inference-time compute paradigm Deadline: April 30, 2025. Details:
1
0
0