Hacker News with Generative AI: Reasoning

Grok 3 Beta – The Age of Reasoning Agents (x.ai)
We are thrilled to unveil an early preview of Grok 3, our most advanced model yet, blending superior reasoning with extensive pretraining knowledge.

Artificial Intelligence, Beta Releases, Reasoning

19 points by zachur 2 days ago | 1 comments

Reasoning models are just LLMs (antirez.com)
It’s not new, but it’s accelerating. People that used to say that LLMs were a fundamentally flawed way to reach any useful reasoning and, in general, to develop any useful tool with some degree of generality, are starting to shuffle the deck, in the hope to look less wrong. They say: “the progresses we are seeing are due to the fact that models like OpenAI o1 or DeepSeek R1 are not just LLMs”.

Artificial Intelligence, Reasoning, Generative AI

66 points by rognjen 12 days ago | 14 comments

LIMO: Less Is More for Reasoning (arxiv.org)
We present a fundamental discovery that challenges our understanding of how complex reasoning emerges in large language models.

Artificial Intelligence, Reasoning, Machine Learning

389 points by trott 12 days ago | 128 comments

Demystifying Long Chain-of-Thought Reasoning in LLMs (arxiv.org)
Scaling inference compute enhances reasoning in large language models (LLMs), with long chains-of-thought (CoTs) enabling strategies like backtracking and error correction.

Generative AI, Artificial Intelligence, Computer Science, Reasoning

11 points by Anon84 13 days ago | 0 comments

Understanding Reasoning LLMs (sebastianraschka.com)
This article describes the four main approaches to building reasoning models, or how we can enhance LLMs with reasoning capabilities. I hope this provides valuable insights and helps you navigate the rapidly evolving literature and hype surrounding this topic.

Artificial Intelligence, Reasoning, Machine Learning

473 points by sebg 15 days ago | 188 comments

Microsoft Phi 4 with R1 Reasoning (huggingface.co)
These LoRA adapters were trained using diverse reasoning datasets that incorporate structured Thought and Solution responses to enhance logical inference.

Generative AI, Artificial Intelligence, Reasoning

6 points by excsn 18 days ago | 2 comments

A Visual Guide to Reasoning LLMs (maartengrootendorst.com)
DeepSeek-R1, OpenAI o3-mini, and Google Gemini 2.0 Flash Thinking are prime examples of how LLMs can be scaled to new heights through “reasoning“ frameworks.

Generative AI, Artificial Intelligence, Reasoning

31 points by headalgorithm 18 days ago | 2 comments

Efficient Reasoning with Hidden Thinking (arxiv.org)
Chain-of-Thought (CoT) reasoning has become a powerful framework for improving complex problem-solving capabilities in Multimodal Large Language Models (MLLMs). However, the verbose nature of textual reasoning introduces significant inefficiencies.

Artificial Intelligence, Reasoning, Computer Science

172 points by fofoz 18 days ago | 43 comments

OpenAI launches o3-mini, its latest 'reasoning' model (techcrunch.com)
OpenAI on Friday launched a new AI “reasoning” model, o3-mini, the newest in the company’s o family of reasoning models.

OpenAI, Artificial Intelligence, Generative AI, Reasoning

19 points by iyn 21 days ago | 2 comments

DeepSeek-R1 at 3,872 tokens / second on a single Nvidia HGX H200 (nvidia.com)
DeepSeek-R1 is an open model with state-of-the-art reasoning capabilities. Instead of offering direct responses, reasoning models like DeepSeek-R1 perform multiple inference passes over a query, conducting chain-of-thought, consensus and search methods to generate the best answer.

Artificial Intelligence, Open Source, Reasoning

13 points by moondistance 22 days ago | 1 comments

Open-R1: an open reproduction of DeepSeek-R1 (huggingface.co)
OpenAI’s o1 model showed that when LLMs are trained to do the same—by using more compute during inference—they get significantly better at solving reasoning tasks like mathematics, coding, and logic.

Open Source, Artificial Intelligence, Reasoning, Computer Science

394 points by jonbaer 24 days ago | 233 comments

Bespoke-Stratos: The unreasonable effectiveness of reasoning distillation (bespokelabs.ai)
We trained Bespoke-Stratos-32B, our reasoning model distilled from DeepSeek-R1 using Berkeley NovaSky’s Sky-T1 data pipeline.

Generative AI, Machine Learning, Reasoning, AI Research

13 points by madiator 27 days ago | 2 comments

Official DeepSeek R1 Now on Ollama (ollama.com)
DeepSeek’s first-generation reasoning models, achieving performance comparable to OpenAI-o1 across math, code, and reasoning tasks.

AI, Reasoning, New Releases

234 points by excerionsforte 32 days ago | 80 comments

Mulberry: Empowering MLLM with o1-like Reasoning (arxiv.org)
In this work, we aim to develop an MLLM that understands and solves questions by learning to create each intermediate step of the reasoning involved till the final answer.

Artificial Intelligence, Reasoning, Research

3 points by Anon84 52 days ago | 0 comments

Coconut by Meta AI – Better LLM Reasoning with Chain of Continuous Thought? (aipapersacademy.com)
Large language models (LLMs) have demonstrated incredible reasoning abilities, penetrating an increasing number of domains in our lives.

Artificial Intelligence, Reasoning

362 points by TaurenHunter 53 days ago | 156 comments

Reimagining mathematics in a world of reasoning machines [video] (youtube.com)

Mathematics, Artificial Intelligence, Reasoning, Video

74 points by kentricon 64 days ago | 41 comments

Are LLMs capable of non-verbal reasoning? (arstechnica.com)
Processing in the "latent space" could help AI with tricky logical questions.

Generative AI, Artificial Intelligence, Reasoning

6 points by Netherland4TW 70 days ago | 0 comments

Phi-4: Microsoft's Newest Small Language Model Specializing in Complex Reasoning (microsoft.com)
Today we are introducing Phi-4, our 14B parameter state-of-the-art small language model (SLM) that excels at complex reasoning in areas such as math, in addition to conventional language processing.

Artificial Intelligence, Microsoft, Reasoning, Software

439 points by lappa 71 days ago | 143 comments

Training LLMs to Reason in a Continuous Latent Space (arxiv.org)
Large language models (LLMs) are restricted to reason in the "language space", where they typically express the reasoning process with a chain-of-thought (CoT) to solve a complex reasoning problem.

Generative AI, Machine Learning, Artificial Intelligence, Reasoning

283 points by omarsar 73 days ago | 114 comments

DeepThought-8B: A small, capable reasoning model (ruliad.co)
Today we're releasing DeepThought-8B, a small, capable AI reasoning model built on LLaMA-3.1 8B.

Artificial Intelligence, New Releases, Reasoning

134 points by AnhTho_FR 86 days ago | 104 comments

The Problem with Reasoners (notion.site)

Reasoning, Philosophy, Artificial Intelligence

32 points by lawrencechen 87 days ago | 2 comments

LLaVA-O1: Let Vision Language Models Reason Step-by-Step (arxiv.org)
Large language models have demonstrated substantial advancements in reasoning capabilities, particularly through inference-time scaling, as illustrated by models such as OpenAI's o1. However, current Vision-Language Models (VLMs) often struggle to perform systematic and structured reasoning, especially when handling complex visual question-answering tasks.

Computer Vision, Reasoning, Artificial Intelligence

177 points by lnyan 95 days ago | 32 comments

Reasonable Person Principle (cs.cmu.edu)
Everyone will be reasonable. Everyone expects everyone else to be reasonable. No one is special. Do not be offended if someone suggests you are not being reasonable.

Reasoning, Social Dynamics, Ethics

75 points by yamrzou 110 days ago | 69 comments

Detecting when LLMs are uncertain (thariq.io)
This post tries to explain the new reasoning techniques developed by XJDR in a new project called Entropix.

Generative AI, Artificial Intelligence, Reasoning

283 points by trq_ 119 days ago | 167 comments

Use Prolog to improve LLM's reasoning (shchegrikovich.substack.com)
On one side, LLMs show unseen capabilities in reasoning, but on the other - reasoning in LLMs is not ideal.

Reasoning, Programming Languages, Artificial Intelligence

378 points by shchegrikovich 131 days ago | 155 comments

Google is working on AI software with human-like reasoning ability (msn.com)

Artificial Intelligence, Google, Reasoning, Technology

4 points by TheAITalk 142 days ago | 0 comments

LLMs still can't reason like humans (freethink.com)
Imagine what would happen if you attempted the following experiment: First, place a washed, fresh tomato and an equally clean carrot on top of a normal kitchen plate. With one hand behind your back, flip the non-stick plate upside-down, inspecting the underside of the plate for marks. Now, slowly turn the plate right-side up and count the number of vegetables remaining on top. How many are on the plate?

Artificial Intelligence, Reasoning, Experimentation

3 points by Brajeshwar 144 days ago | 0 comments

Deductive Verification for Chain-of-Thought Reasoning in LLMs (arxiv.org)
Large Language Models (LLMs) significantly benefit from Chain-of-Thought (CoT) prompting in performing various reasoning tasks.

Generative AI, Artificial Intelligence, Reasoning, Computer Science

80 points by smooke 164 days ago | 20 comments

Inductive or deductive? Rethinking the fundamental reasoning abilities of LLMs (arxiv.org)