Reasoning models, like OpenAI o1 and o3-mini, are new large language models trained with reinforcement learning to perform complex reasoning. Reasoning models think before they answer, producing a long internal chain of thought (test-time compute before responding to the user. Reasoning models excel in complex problem solving, coding, scientific reasoning, and multi-step planning for agentic workflows.1
목차 목차 관련 글 The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity “This setup enables the analysis of not only final answers but also the internal reasoning traces, offering insi…” Sparks of artificial general intelligence: Early experiments with GPT-4 이 글을 인용: “한편 GPT4 이후 추론 모델들이 나오면서 이 문제를 약간 개선.” Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens “With this lens, we dissect CoT reasoning via three dimensions: task, length, and format.” Reliable Reasoning Beyond Natural Language “To address this, we propose a neurosymbolic approach that prompts LLMs to extract and encode all relevant infor…” Large Language Models are zero-shot reasoners “Notably, Chain-of-thought, a recent technique for eliciting complex multi-step reasoning through step-by-step a…” Large language models are human-level prompt engineers “모델의 답변이 얼마나 신뢰할만한지 평가하기 위해, 법/건강/금융/정치 등 38개 분야의 817개 질문으로 구성된 데이터셋.” 경험의 시대에 오신 걸 환영합니다 “Planning and Reasoning: 에이전트는 인간적 관점에서의 추론을 하는 게 아니라 스스로의 경험에 대해 계획하고 추론할 것이다.” Tensor Logic: The Language of AI “Most importantly, tensor logic makes new directions possible, such as sound reasoning in embedding space.” Prompting Science Report 2: The Decreasing Value of Chain of Thought in Prompting “CoT is a widely adopted method for improving reasoning tasks, however, our findings reveal a more nuanced pictu…” Nested Learning: The Illusion of Deep Learning Architectures “Combining our self-modifying sequence model with the continuum memory system, we present a learning module, cal…” MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark “In the age of large-scale language models, benchmarks like the Massive Multitask Language Understanding (MMLU) …” 개념 모델 관련 절: 개념 모델의 설계 vs. 커뮤니케이션 Thinking Beyond Tokens: From Brain-Inspired Intelligence to Cognitive Foundations for AGI and its Societal Impact “We analyze the architectural and cognitive foundations of general intelligence, highlighting the role of modula…” Language Models Know More Than They Show: Exploring Hallucinations From the Model's Viewpoint “Large language models (LLMs) often produce errors, including factual inaccuracies, biases, and reasoning failur…” Some evidence on the cognitive benefits of learning to code “Part of the reasoning is that learning to code may ultimately aid students' learning and acquiring of skills in…” Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models “Large language model (LLM) applications such as agents and domain-specific reasoning increasingly rely on conte…” The cognitive benefits of learning computer programming: A meta-analysis of transfer effects “Larger gains in specific skills closely tied to coding (creative problem-solving, mathematical reasoning, etc.).” Vibe Coding vs. Agentic Coding: Fundamentals and Practical Implications of Agentic AI “While both leverage large language models (LLMs), they differ fundamentally in autonomy, architectural design, …” Let Me Speak Freely? A Study on the Impact of Format Restrictions on Performance of LLMs “Surprisingly, we observe a significant decline in LLMs reasoning abilities under format restrictions.” Introduction to Machine Learning 관련 절: Definition of [[Machine learning|ML]]