Skip to content

#Reasoning

0 today

Nov 13

Friday
  1. OpenAI’s data suggests that frontier models are hiding their thoughts because developers optimize for capability and compute

    OpenAI's safety team systematically investigated why the chains of thought in frontier models like GPT-6 Astra are increasingly evading oversight. Four training steps drive the erosion, including architecture choices and reinforcement learning. The analysis shows that the loss of transparency is not a law of nature but also reflects tradeoffs around capability and cost.

Nov 9

Monday

Nov 8

Sunday

Nov 4

Wednesday

Nov 3

Tuesday

Oct 31

Saturday
  1. dbreunig/nesformer

    dbreunig/nesformer Inspired by qwenputer by Divyansh Agrawal, Drew Breunig baked a working emulator for a Nintendo Entertainment System 6502 CPU into the weights of Qwen3.5-4B. You can now play NES ROMs such as Super Mario Bros (very slowly) running inside Qwen!

Oct 28

Wednesday

Oct 22

Thursday

Oct 21

Wednesday

Oct 17

Saturday

Oct 16

Friday

Oct 15

Thursday
  1. GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece

    On the TheoPhysEval benchmark, GPT-6-Astra solves 49 of 50 unpublished theoretical physics calculations as accurately as leading specialists. In some cases where the agent seemed to fail, the experts' own reference solutions turned out to contain errors. But the researchers say physics hasn't had a breakthrough like the ones in mathematics, because the agents lack "taste."

Oct 13

Tuesday

Oct 12

Monday

Oct 7

Wednesday

Oct 6

Tuesday

Oct 2

Friday

Oct 1

Thursday

Sep 16

Wednesday

Sep 8

Tuesday

Sep 2

Wednesday

Aug 25

Tuesday

Aug 17

Monday

Aug 13

Thursday

Jul 31

Friday

Jul 29

Wednesday
  1. From CUDA to MLX: How K-Search Brings Decades of Kernel Expertise to Apple Silicon

    Figure 1: CUDA-to-MLX optimization translation map. CUDA optimization knowledge can be translated into architecture-native MLX strategies rather than copied instruction-for-instruction. We face a new epoch in computing. Hardware is changing rapidly — not just faster GPUs, but a growing range of chips from different vendors, each with its own architecture and often tailored to specific AI workloads.

Jul 26

Sunday

Jun 25

Thursday
  1. Thinking to recall: How reasoning unlocks parametric knowledge in LLMs

    A Google Research paper at COLM 2026 finds reasoning traces unlock factual knowledge LLMs otherwise cannot recall, even for simple single-hop questions. Tests on Gemini-2.5 Flash and Pro and Qwen3-32B identify two mechanisms: extra tokens act as a 'computation buffer', and 'factual priming' produces related facts to semantically prepare the correct answer. Self-generated intermediate facts can also introduce hallucination risks.

Apr 20

Monday
  1. Gradient-based Planning for World Models at Longer Horizons

    GRASP is a new gradient-based planner for learned dynamics (a “world model”) that makes long-horizon planning practical by (1) lifting the trajectory into virtual states so optimization is parallel across time, (2) adding stochasticity directly to the state iterates for exploration, and (3) reshaping gradients so actions get clean signals while we avoid brittle “state-input” gradients through high-dimensional vision models.

Apr 16

Thursday

Mar 25

Wednesday

Mar 13

Friday
  1. Identifying Interactions at Scale for LLMs

    --> Understanding the behavior of complex machine learning systems, particularly Large Language Models (LLMs), is a critical challenge in modern artificial intelligence. Interpretability research aims to make the decision-making process more transparent to model builders and impacted humans, a step toward safer and more trustworthy AI.

Nov 1

Saturday
  1. RL without TD learning

    In this post, I’ll introduce a reinforcement learning (RL) algorithm based on an “alternative” paradigm: divide and conquer . Unlike traditional methods, this algorithm is not based on temporal difference (TD) learning (which has scalability challenges ), and scales well to long-horizon tasks. We can do Reinforcement Learning (RL) based on divide and conquer, instead of temporal difference (TD) learning.