Reinforcement learning (RL) post-training improves the reasoning capabilities of large language models, but its cost is dominated by autoregressive rollout generation, which can exceed 70% of training time. Existing speculative-decoding methods are poorly suited to this setting: model-based drafters consume scarce GPU...
N. Pandey, Jangseon Park, Jason Kong et al.· 0 citations
Hyperdimensional computing (HDC) is attractive for efficient and robust learning, but conventional inference still encodes every query independently, repeatedly paying the cost of high-dimensional projection. We introduce SupHDC, a new inference paradigm that processes multiple queries through a shared encoding computa...
Quan-Ling Zhao, N. Pandey, Ye Tian et al.· 0 citations
LIMBO is introduced, the first online framework to the authors' knowledge that treats memory as a controllable inference-time resource and jointly optimizes memory strategy and inference budget for each incoming task, and learns this allocation online in a single pass.
Siddharth Sharma, N. Pandey, Onat Gungor et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.