Search: “reasoning”

This is brilliant. Recommended reading. It's a great example of how you could potentially build more effective harnesses using System One and System Two models. It explores Jev's potential as a fuzzy

AI grad student here. With all the recent interest in Jev, I thought I would share something I built la few months ago that brings ML models and LLMs together in a different way than Jev does for diff

ggml-cpu: tiled mul_mat for k-quants (#27851) Added tiled mul_mat. For each mul_mat_one_chunk, quants are unpacked into (max) 256x256 tiles of int8, one routine per quent. Then microkernel computes 16

About a month ago I was having FOMO and was going to spend coin I don't really have on new graphics cards. Instead of doing that though I decided to spend the money on a new motherboard+cpu and try to

Sometimes it’s easy to forget that this sub and others like it are probably the extreme minority when it comes to this hobby. Most people, I would think, don’t use or can’t afford one good GPU, let al

i don’t get it fully but this looks cool We achieved a breakthrough at BTL. AI models are very linear. They think in one direction at a time, and when they're wrong they scramble and start again. For

I've been playing around with Ling 3.0 Tiny, which is an 8 billion parameter model (MoE, 1B active). And I've had a lot of poignant thoughts as a result. Just for fun, I got it running with llama.cpp

This is huge. Xiaomi just open-sourced 7k+ reinforcement learning task environments used to train MiMo, covering code, cybersecurity, general tool use, visual web development, and music-related tasks.

No 5hr limits for Plus, strong Sol in chat that could actually work on problems for a long time. That’s what made me subscribe. Now the product feels weaker somehow, it’s not just the 5hr limits back

Worst thing about this is we have no idea if GPT-6 was specifically post trained on Nethack or not. Zero transparency about the training data and environments. You can't independently evaluate the sig

arXiv:2609.30205v1 Announce Type: new Abstract: Large language model (LLM)-based clinical assistants are increasingly being integrated into electronic health record (EHR) systems, transforming how cli

arXiv:2604.14590v3 Announce Type: replace-cross Abstract: In modern data-streaming systems, alongside traditional programs, a new type of entity has emerged that can interact with streaming data: AI a

arXiv:2609.30184v1 Announce Type: new Abstract: U.S. employment-discrimination complaints describe complex event sequences that are not explicitly captured by lexical or embedding-based representation

Banger paper from Microsoft and colleagues. We talk about human taste being important in this AI era. But for recursive self-improvement, an agent's taste also matters. The big question is: How often

Banger report on building faster memory for AI Agents. If your agent's memory layer is slow, this design is worth a look. Great weekend read. It proposes a new agentic memory architecture, Jev-Mem, in

arXiv:2609.29519v1 Announce Type: new Abstract: Mental health disorders affect hundreds of millions of people around the world, yet access to professional counseling remains severely limited. AI-power

arXiv:2609.29735v1 Announce Type: new Abstract: Long-horizon tasks require preserving and later recovering cross-session evidence under a bounded, query-blind memory budget. Existing compression can d

arXiv:2609.29828v1 Announce Type: new Abstract: We present ChunkRank, an open-source Python library that derives chunk boundaries from a target model's tokenizer and context window, and selects an ans

arXiv:2609.28984v1 Announce Type: cross Abstract: Cross-embodiment learning has shown that a single model, such as a vision-language-action (VLA) model, can learn state representations and manipulatio

arXiv:2609.29669v1 Announce Type: cross Abstract: Robot learning now advances along two tracks that rarely meet. On one side, direct Vision-Language-Action (VLA) policies map observations to actions a

arXiv:2609.29508v1 Announce Type: new Abstract: Large language model (LLM) agents may perform well on isolated tasks yet drift into inconsistency over extended interaction. We evaluate temporal consis

arXiv:2609.28949v1 Announce Type: new Abstract: Few-shot segmentation (FSS) aims to segment unseen object categories with a few (e.g., one or five) labeled examples, enabling efficient adaptation to n

arXiv:2609.29692v1 Announce Type: new Abstract: Fair allocation of scarce, indivisible resources is an important challenge in many societal problems. While there are several formal theories of fairnes

arXiv:2501.09223v3 Announce Type: replace-cross Abstract: This is a book about large language models. As indicated by the title, it primarily focuses on foundational concepts rather than comprehensive

arXiv:2609.29601v1 Announce Type: cross Abstract: End-to-end vision-language models (VLMs) bind visual competence to the scale of their language model: as the language model shrinks, perception and re

arXiv:2609.29091v1 Announce Type: new Abstract: Recent advances in agentic systems have substantially enhanced the long-horizon capability of embodied manipulation. However, many existing frameworks s

arXiv:2609.30147v1 Announce Type: new Abstract: Large Language Models (LLMs) typically exhibit a performance profile where reliability degrades as task complexity increases. We address the challenge o

arXiv:2609.26760v2 Announce Type: replace Abstract: Large language model (LLM) agents often handle streams of related tasks, yet standard harnesses repeatedly ask the model to reconstruct the same con

The jump from Qwen3 Coder 30B A3B to current-day Qwen 3.6 35B A3B is crazy, especially with all the fine-tunes, and that was around 6 months (I didn't care for local AI back then, or AI at all, apart

Every rent-vs-buy thread I read has confident people on both sides, but not many actually show the numbers. So I finally ran the numbers for our own decision. Posting the working here in case it is us

arXiv:2609.30217v1 Announce Type: cross Abstract: A central concern in AI safety is that agents may treat oversight as an obstacle when it conflicts with completing their goals. We study instrumental

Zvi Mowshowitz / Don't Worry About the Vase: Jensen Huang tells Ezra Klein that AI is just software whose existential risks are overstated, yet his standards would shut down OpenAI and 10x safety spen

arXiv:2609.30186v1 Announce Type: new Abstract: Vision-language models (VLMs) have become a common foundation for autonomous mobile GUI agents, but most existing systems rely on the VLM for both plann

I've been building a small decision model for agent loops: gates, routers, 'should I ask the user or just act' checks. It's out now as Mica v0.1 4B (Apache-2.0). What it does You give it a state, a qu

arXiv:2609.29913v1 Announce Type: new Abstract: Many-shot in-context learning (ICL) enables large language models (LLMs) to adapt to complex tasks by conditioning on thousands of demonstration example

arXiv:2602.08329v2 Announce Type: replace-cross Abstract: A core bottleneck in large language model (LLM) inference is the cost of attending over the ever-growing key-value (KV) cache. Although near-o

arXiv:2609.29317v1 Announce Type: cross Abstract: Time series analysis is fundamental in domains such as finance, healthcare, and meteorology. Real-world time series often exhibit multiscale character

arXiv:2609.29245v1 Announce Type: cross Abstract: Given a large corpus, the questions one might ask can vary -- from 'When was the first human heart transplant?' to 'What are all the contradictory cla

Nope. It's still true. Where is your domestic robot? Where is your Level-5 self-driving car? Where is your robot car that can learn to drive in a few hours of practice like any 17 year old? Where is y

arXiv:2603.11351v2 Announce Type: replace-cross Abstract: In dynamic open-world environments, autonomous agents often encounter novelties that hinder their ability to find plans to achieve their goals

arXiv:2609.28854v1 Announce Type: cross Abstract: Language-model agents increasingly answer questions over customer-relationship management (CRM) records, such as whether to qualify a sales lead. We i

arXiv:2609.29233v1 Announce Type: cross Abstract: We find that language models can transfer capabilities through task-unrelated text. Post-training typically improves language models using task-specif

arXiv:2609.29718v1 Announce Type: cross Abstract: Coding agents are beginning to act in the visual world. They now build webpages, GUIs, games, 3D scenes, diagrams, and documents. Success in such visu

Prefill step size affects both PP performance and your drafter fetching logits for MTP (affects dflash as well, mlx-vlm needs to be patched slightly to support chunked prefill for dflash). It also nee

arXiv:2609.29892v1 Announce Type: new Abstract: The rapid progression of large language models is extending AI from passive content generation into the active workflows of engineering and scientific d

arXiv:2609.30100v1 Announce Type: cross Abstract: Recent advances in paraphrase detection reveal a fundamental trade-off: large language models achieve high accuracy but require high computation, whil