Dietrich Stein (@pixelsort)

Anthropic가 지난달 @deepseek_ai 등 일부 연구실이 자사 모델의 능력을 '도용'했다고 폭로했고, 결과적으로 해당 모델들의 체인오브Thought(Chain of Thought) 추적(trace)이 더 이상 보이지 않게 되었다는 내용입니다. 작성자는 안타까워하면서도 구글의 Gemini는 여전히 CoT를 제공한다고 언급하고 있습니다.

https://x.com/pixelsort/status/2032530587072741710

#anthropic #deepseek #chainofthought #gemini #aisafety

Dietrich Stein (@pixelsort) on X

Last month, @AnthropicAI revealed that @deepseek_ai and other labs have been "stealing" their capabilities. Consequently, we can no longer see the Chain of Thought traces in their models. I'm sympathetic, but saddened. At least @Gemini still has them. https://t.co/eKD4Vwil2H

X (formerly Twitter)

New research shows TensorRT Edge‑LLM can run chain‑of‑thought reasoning directly on devices, boosting physical AI tasks like autonomous‑vehicle perception and MATH500 benchmarks. Efficient, on‑device inference means smarter, safer robots without cloud latency. Dive into the details of this breakthrough for on‑device language models. #TensorRT #EdgeLLM #ChainOfThought #PhysicalAI

🔗 https://aidailypost.com/news/tensorrt-edgellm-enables-efficient-chainofthought-processing-physical

fly51fly (@fly51fly)

2026년 논문 'Reasoning Models Struggle to Control their Chains of Thought'는 추론 모델들이 자신의 체인오브소트(Chain of Thought)를 제어하는 데 어려움을 보인다는 분석을 제시한다. C Yueh-Han, R McCarthy, B W. Lee, H He 등(NYU·UCL·OpenAI 소속)이 공동저자로 arXiv에 공개됨.

https://x.com/fly51fly/status/2031126438292894184

#reasoning #chainofthought #airesearch #modelbehavior

fly51fly (@fly51fly) on X

[AI] Reasoning Models Struggle to Control their Chains of Thought C Yueh-Han, R McCarthy, B W. Lee, H He… [NYU & UCL & OpenAI] (2026) https://t.co/kR3dSHR50x

X (formerly Twitter)

Chain of Thought bleibt sichtbar.

Aktuelle Reasoning-Modelle können ihre internen Rechenschritte nicht vor Monitoring-Systemen verbergen. Studien zeigen, dass Versuche zur Verschleierung – etwa durch Keyword-Vermeidung – meist fehlschlagen. Besonders bei langen Rechenketten bricht die Kontrolle über die eigene Ausgabe zusammen. Die Analyse der Zwischenschritte bleibt damit ein valider Weg für Sicherheitschecks.

#OpenAI #KISicherheit #ChainOfThought
https://www.all-ai.de/news/beitrage2026/openai-studie-neu-luegen

Moderne KI-Systeme können ihre echten Absichten nicht verschleiern

Die internen Denkprozesse von Reasoning-Modellen bleiben für Sicherheitsmechanismen lesbar. Manipulationsversuche schlagen derzeit meistens fehl.

All-AI.de

OpenAI (@OpenAI)

새로운 평가 도구와 연구 논문 공개: Chain-of-Thought(이하 CoT) 제어성 평가 스위트를 발표하며, 연구 결과 GPT-5.4 Thinking은 자신의 추론을 은폐하는 능력이 낮아 CoT 모니터링이 여전히 유용한 안전 도구임을 시사한다고 보고했다.

https://x.com/OpenAI/status/2029650046002811280

#chainofthought #evaluation #safety #gpt5.4

OpenAI (@OpenAI) on X

We're publishing a new evaluation suite and research paper on Chain-of-Thought (CoT) Controllability. We find that GPT-5.4 Thinking shows low ability to obscure its reasoning—suggesting CoT monitoring remains a useful safety tool. https://t.co/isZkNkPXZm

X (formerly Twitter)

Tencent HY (@TencentHunyuan)

Tencent HY 팀의 25편 이상 논문이 CVPR 2026에 채택되었다고 발표했습니다. 이번 작업은 이미지 편집 분야에서 Chain-of-Thought(Cot) 추론 및 도구 호출을 통한 정밀·제어 가능한 편집, 비디오 생성 분야에서는 사후 학습 기반 강화학습(post-training RL)과 오디오·비디오 통합(unified audio-video) 등을 포함합니다.

https://x.com/TencentHunyuan/status/2029123001913098645

#cvpr #computervision #imageediting #videogeneration #chainofthought

Tencent HY (@TencentHunyuan) on X

More than 25 papers from Tencent HY team have been accepted to @CVPR 2026. This year, our work spans: 🖼️ Image Editing: Chain-of-Thought (CoT) reasoning and tool-calling for fine-grained, controllable editing. 🎥 Video Generation: Post-training RL, unified audio-video

X (formerly Twitter)

fly51fly (@fly51fly)

Effective Reasoning Chains 논문은 체인 기반 추론(reasoning chains)이 모델 내부 표현의 내재적 차원(intrinsic dimensionality)을 낮춘다는 발견을 보고합니다. A. Prasad, M. Joshi, K. Lee, M. Bansal( Google DeepMind & UNC Chapel Hill )의 이론·실험은 체인 오브 사고의 구조적 이유와 모델 설계·효율성에 대한 시사점을 제공합니다.

https://x.com/fly51fly/status/2021699501342241171

#reasoning #chainofthought #representationlearning #arxiv

fly51fly (@fly51fly) on X

[CL] Effective Reasoning Chains Reduce Intrinsic Dimensionality A Prasad, M Joshi, K Lee, M Bansal... [Google DeepMind & UNC Chapel Hill] (2026) https://t.co/rNsRvQqkz4

X (formerly Twitter)
#EricJang argues that #AImodels can now genuinely think and code. Using #ClaudeCode, he demonstrates #automatedresearch workflows, traces reasoning’s evolution from #ChainofThought to #DeepSeekR1, and predicts massive demand for inference compute. #Codingagents will fundamentally transform #softwareengineering, #research, and #militarystrategy - “the rocks can think now.“​​​​​​​​​​​​​​​​ https://evjang.com/2026/02/04/rocks.html?eicker.news #tech #media #news
As Rocks May Think

You are viewing the mobile version of this page. This content is best viewed on a desktop.

Eric Jang

New research shows DeepSeek-R1 and QwQ-3 develop distinct personalities that boost chain-of-thought reasoning, hinting at a future where societies of thought among LLMs improve problem solving. Open-source enthusiasts, see how personality diversity reshapes AI reasoning! #DeepSeekR1 #QwQ32B #ChainOfThought #PersonalityDiversity

🔗 https://aidailypost.com/news/deepseekr1-qwq3-exhibit-competing-personalities-that-improve-reasoning

fly51fly (@fly51fly)

논문 'Reasoning about Reasoning: BAPO Bounds on Chain-of-Thought Token Complexity in LLMs' (2026) 발표 알림: 저자 K. Tomlinson, T. Schnabel, A. Swaminathan, J. Neville 등이며 Microsoft Research와 Netflix 소속 연구자들이 참여. 체인-of-사고(Chain-of-Thought)의 토큰 복잡도에 대한 BAPO 경계 이론을 제시하는 연구입니다.

https://x.com/fly51fly/status/2019163030630855088

#chainofthought #llm #theory #research

fly51fly (@fly51fly) on X

[LG] Reasoning about Reasoning: BAPO Bounds on Chain-of-Thought Token Complexity in LLMs K Tomlinson, T Schnabel, A Swaminathan, J Neville [Microsoft Research & Netflix] (2026) https://t.co/C0lUKMa0yj

X (formerly Twitter)