Upcoming EBR paper claims 345× fewer FLOPs than looped reasoning at 99.5%+ accuracy
YesThisIsLion · x · 2026-11-03
@AlexiGlad teased a paper dropping this week on Energy-Based Reasoning (EBR), claiming "looped reasoning is completely overrated." The attached chart shows EBR reaching 99.5–99.9% accuracy at roughly 0.5 TFLOPs per puzzle, while looped-reasoning methods TRM and GRAM need 1–50 TFLOPs and EqR 200 TFLOPs — a claimed 345× FLOPs reduction vs EqR. Numbers are unverified until the paper lands, but if it holds up it's a serious challenge to iterative-refinement reasoning.
More from Research
- Quiz: Can You Diagnose All Four Training Scenarios From Their Loss Curves? — cneuralnetwork · 2026-11-03
- NVIDIA, Tsinghua, CMU, Berkeley and Princeton launch StreamEval, first real-time interactive streaming video benchmark — JaynitMakwana · 2026-11-03
- MIT: AI edits 0.016% of another model's weights, lifting MATH-500 from 7.9% to 48.1% — rohanpaul_ai · 2026-11-03
- Schmidhuber: AI weight editing is 'old hat' from his 1990s fast weight programmers — SchmidhuberAI · 2026-11-03
- UCL's TAL lets LLM agents learn from tutorial videos with no parameter updates, +113% on VidLearn — UniversityCollegeLondon · 2026-11-03
- SIMIT: vision-language models self-improve at test time by imagining their own practice data — dtudk · 2026-11-03