Looped Flows Paper Hits 58.8% on ARC-AGI-1 With Recurrent Denoising Reasoning
MiniMax_AI · x · 2026-09-14
A new arXiv paper, "Thinking with Looped Flows," combines flow/diffusion-style denoising with recurrent reasoning. Instead of backpropagating through one or two recurrent updates (BPTT), it trains the recurrence with local denoising objectives — progressively decreasing noise levels and shared noise encourage hidden states that transfer useful computation over time. Inference integrates a probability-flow velocity parameterized by the learned denoiser, so more compute at test time means a finer temporal grid, and different initial noise samples yield multiple valid predictions. Across six reasoning benchmarks, looped flows beat prior looped-model SOTA, reaching 58.8% on ARC-AGI-1 and 12.2% on ARC-AGI-2.
Related event: Looped Flows Paper Combines Diffusion Denoising with Recurrent Reasoning(2 posts)→
More from Research
- Tencent's retrieval models dominate Hugging Face trends with 4 entries — tomaarsen · 2026-09-14
- TII-254 McEliece key recovery challenge solved with just 27.2 GPU-hours — jedisct1 · 2026-09-14
- Diversity-Aware Skill Routing Uses DPP to Cut Redundant LLM Agent Skill Picks — Wang Wei · 2026-09-14
- Contextual Bandit Algorithms Route Prompts to LLM Experts with Sublinear Regret — Wang Wei · 2026-09-14
- Heidelberg Laureate Forum launches daily podcast, first episode features Fintzen and Hoefler — thoefler · 2026-09-14
- WebMCP browser tools cut tokens 52% on one task but increase them on another, DeepDeck experiment finds — j032 · 2026-09-14