Monday, October 26, 2026i10X News · Section desk7 stories
Studies
Reasoning
Reasoning, planning, math and chain-of-thought research
FilterCompanies & models
-
Randomly Evicting KV Cache Matches Smart Selectors — and Runs 32–43% Faster
Salesforce’s Random Attention keeps the prompt and randomly drops reasoning-trace tokens per head. Across 4 models and 6 tasks it matches the best scorer while boosting vLLM throughput 32–43%.
By
Editor's picks
- Randomly Evicting KV Cache Matches Smart Selectors — and Runs 32–43% Faster
- A 150M Model Hits 29.5% on ARC-AGI-1 for Under a Tenth of a Cent
- RL for Reasoning Wastes Gradient on Formatting Tokens. DelTA Fixes the Credit
- Distill the RL Direction in Hidden Space, Not Noisy Output Rewards
- A 3B Model Hits 94.3 on AIME26 and 80.2 on LiveCodeBench
More stories
-
RL Post-Training Makes Agents More Accurate — and Less Able to Explore
Across 14 base/post-trained pairs on 3 agentic benches, base models often beat RL successors at pass@K. On WebShop, gemma-4-31B’s “pass given compute” shrinks from 87.6% to…
-
Turn Off Thinking Mode and Models Still Reason 99.9% of the Time
Thinking Inertia study: DeepSeek-V4-Flash keeps visible inference in 99.9% of open-ended answers with thinking disabled. Strict answer-only prompts raise Empty-Thinking Rate to ~40% but cut accuracy…
· RSS