Search: “llm”

ReSight-SMC: Two-Stage Power Sampling via Island SMC with Visual Scouts

model-releasesarxiv-cs-cv

arXiv:2609.34905v1 Announce Type: new Abstract: Power sampling has emerged as a training-free approach to LLM reasoning, eliciting capabilities comparable to reinforcement learning by sharpening the m

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

RubiCap: Rubric-Guided Reinforcement Learning for Dense Image Captioning

DGX agent

arXiv:2603.09160v2 Announce Type: replace Abstract: Dense image captioning is critical for vision-language pretraining and for text-to-image generation, but scaling expert-quality annotations is prohi

safetyarxiv-cs-cv

29 Sept 2026

Tutorials

SeLMRoute: Probabilistic Semantic Evidence for Large Language Model Routing

arXiv:2609.34736v1 Announce Type: cross Abstract: Large language model (LLM) routing aims to select the most suitable model for each incoming query. Most existing routers learn this decision directly

tutorialsarxiv-cs-lg

Sherry's 3:4 ternary format (1.375 bits per weight) running on WebGPU: a 1.6 MB model that plays Connect Four as well as its 7.8 MB int8 version

model-releasesr-localllama

Not an LLM, but the ternary findings should carry over, and we hadn't seen Sherry-style 3:4 weights run in a browser before. Disclosure: this is our work at Precisit, everything is MIT. What it is A 7

Towards Autonomous and Auditable Medical Imaging Model Development

arXiv:2607.10522v2 Announce Type: replace Abstract: Large language model (LLM) agents are beginning to automate machine learning engineering (MLE) by coupling planning, code execution, debugging, and

agentsarxiv-cs-cv

Understanding and Exploiting Anisotropy in Post-Training

model-releasesarxiv-cs-lg

arXiv:2609.32792v1 Announce Type: new Abstract: LLM post-training combines supervised fine-tuning (SFT), a mode-covering forward-KL objective, with reinforcement learning (RL), a mode-seeking reverse-

Unveiling Implicit Advantage Symmetry: Why GRPO Struggles with Exploration and Difficulty Adaptation

arXiv:2602.05548v4 Announce Type: replace Abstract: Reinforcement Learning with Verifiable Rewards (RLVR), particularly GRPO, has become the standard for eliciting LLM reasoning. However, its efficien

researcharxiv-cs-lg

WirelessMathBench-XL: An Auditable Benchmark for Wireless Mathematical Reasoning

model-releasesarxiv-cs-lg

arXiv:2509.23219v2 Announce Type: replace Abstract: Technical-domain benchmarks constructed from arXiv papers can overlap the same public text used in LLM pretraining. Auditing this risk at training-c

A Benchmarking Framework for Context-aware XR Interfaces

arXiv:2609.30466v1 Announce Type: cross Abstract: Everyday Extended Reality (XR) systems aim to provide context-aware access to the right functionalities at the right time and place, with minimal manu

A Framework for Identifying, Categorizing, and Explaining Bias in AI-Generated Code

arXiv:2609.30642v1 Announce Type: cross Abstract: As Large Language Models (LLMs) become integrated into software development workflows, concerns regarding unintentional biases in AI-generated code. A

Actively Resolving Contextual Uncertainty for Underspecified Tasks in Natural Language

arXiv:2609.30428v1 Announce Type: cross Abstract: Foundation models provide robots with the ability to interpret natural language and reason about environmental context, yet most language-conditioned

America is starting to be World Model-pilled

model-releasescristobal-valenzuela--x

America is starting to be World Model-pilled Heard a crazy stat the other day. Video generation represents around 70% of the share of AI token consumption in China. A combination of short form content

CacheReforge: Bounded Recovery for Stale KV Caches under Evolving Adapters

arXiv:2609.30884v1 Announce Type: new Abstract: Large language models rely on KV caching to reduce repeated prefill computation in long context and interactive applications. As lightweight adapters ev

researcharxiv-cs-lg

CG-Probes: Recovering Guardrail Directions from Patient Query Embeddings

arXiv:2609.31062v1 Announce Type: new Abstract: Patient-facing AI assistants promise valuable support to patients, but incoming queries can pose medical risks. To create guardrails, we work with oncol

safetyarxiv-cs-cl

Hardware

EAServe: Encode-Aware Disaggregated Serving for Multimodal Large Language Models

arXiv:2609.31551v1 Announce Type: cross Abstract: Disaggregating the two stages, Prefill and Decode, onto separate GPU pools is now a standard optimization for (text-only) LLM serving. However, multim

hardwarearxiv-cs-lg

Efficient Safety Benchmarking via Item Response Theory

arXiv:2606.20626v2 Announce Type: replace-cross Abstract: Safety benchmarks for language models are typically evaluated using static paradigms that treat all items as equally informative for all model

Evaluating Sycophancy in Chinese Large Language Models on Factual Questions Derived from Online Search Queries

model-releasesarxiv-cs-cl

arXiv:2609.30986v1 Announce Type: new Abstract: As large language models increasingly mediate information access, factually accurate and independent answers are critical. However, these models can exh

FTB Graph: Determining and Validating First-token Broadcasters and Language-Identity Head Circuits in Multilingual Language Models

arXiv:2609.30954v1 Announce Type: new Abstract: Large language models operating in multilingual contexts must resolve target response languages early in generation, yet the causal circuitry governing

Applications

REALMS: An AI-Assistant Conversational System for Real-Time Exact Audience Sizing over High-Dimensional Nested Profiles

arXiv:2609.30547v1 Announce Type: new Abstract: Audience sizing is a critical component of digital marketing. It enables precise resource allocation, campaign planning, and performance optimization. T

applicationsarxiv-cs-cl

Resource-Optimized and Energy-Aware Agentic AI Framework Anchored on Blockchain for Secure Software Supply Chains

arXiv:2609.31282v1 Announce Type: cross Abstract: This paper proposes a blockchain-backed agentic security framework designed to safeguard the complete software development lifecycle (SDLC) while also

SkillEvoReg: Regularizing Agent Skill Evolution Against Overfitting

arXiv:2609.30861v1 Announce Type: new Abstract: Language-model agents increasingly improve by converting execution experience into reusable external skills. Yet repeated skill updates form a learning

agentsarxiv-cs-ai

Softmax Reparameterization for Output-Head Quantization

arXiv:2609.31291v1 Announce Type: cross Abstract: Large vocabularies make output heads a substantial inference cost in small language models. We propose softmax reparameterization, a post-training met

Strategically Diverse Sampling for Self-Training

arXiv:2609.31571v1 Announce Type: new Abstract: Many LLM training and inference methods, including RL and test-time scaling, depend on repeated sampling, but benefit only when the responses meaningful

researcharxiv-cs-cl

Subjects, Not Authors: The Authorship Hazard in Agentic Dataspaces

arXiv:2609.30614v1 Announce Type: cross Abstract: Dataspace connectors decide whether a transfer may occur, not what the transferred value contains, tolerable for contracted applications, not for LLM

The KV Cache Is the New Memory Wall

model-releasesarxiv-cs-lg

arXiv:2609.30854v1 Announce Type: cross Abstract: Autoregressive LLM inference at long context is bounded by memory bandwidth, not arithmetic throughput, and the binding resource shifts from model wei

User Model Extraction via Belief Self-Distillation

arXiv:2609.31603v1 Announce Type: cross Abstract: Large language models (LLMs) implicitly infer attributes of their users and adapt their behavior accordingly, yet these beliefs remain difficult to in

safetyarxiv-cs-cl

Warned alike, AI agents avoid the less-crowded road while people take it

arXiv:2609.30883v1 Announce Type: cross Abstract: AI agents built on a few shared models increasingly act for many people. A shared forecast about others can align their choices and change how scarce

agentsarxiv-cs-ai

Applications

Werracle: Sub-Cent Intra-Block AI Reflex Oracles and Flash-Loan Circuit Breakers for EVM Smart Contracts

arXiv:2609.30719v1 Announce Type: cross Abstract: Contemporary on-chain artificial intelligence (AI) encounters an intractable Von Neumann memory and latency wall. Storing static floating-point neural

applicationsarxiv-cs-ai

Which Influence Are We Estimating? The Role of Counterfactual Specifications in Data Attribution

arXiv:2609.31214v1 Announce Type: new Abstract: Estimating the influence of training examples on model behavior is essential for data debugging, valuation, and attribution. Existing influence estimato

A little Ember-1 tl;dr. Seems like a great model! (Was recently asked on a podcast, given $ xx million, what's the best way to develop a fro…

researchsebastian-raschka--x

A little Ember-1 tl;dr. Seems like a great model! (Was recently asked on a podcast, given $ xx million, what's the best way to develop a frontier LLM today? My recommendation was: start with an existi

Cool paper on catching alignment failures with Jev. The ideas is to ask Jev one generic yes/no question about a model's response, and use it…

Cool paper on catching alignment failures with Jev. The ideas is to ask Jev one generic yes/no question about a model's response, and use its probability as a score. With no extra training, that score

model-releasesdair-ai--x

Local Ai

Hardware Roofline Inference Calculator

Hello everyone, I made a calculator for the theoretical HW roofline for decoding / prefill based on several parameters (LLM model architecture, quants, GPU, Memory, ..). It still a theoretical bound,

local-air-localllama

Zer0Fit - Zero-shot predictions, classifications, and regressions using Google ML research models running locally as a dockerized MCP

model-releasesr-localllama

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

i don’t get it fully but this looks cool

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

agentsyohei-nakajima--x

26 Sept 2026

A Living Benchmark for Information Retrieval from Electronic Health Records

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

Local Ai

AlphaDiverse: Post-Training Local Quantitative Research Agents for Diverse Exploration in Alpha Factor Mining

arXiv:2609.29014v1 Announce Type: new Abstract: Large language model (LLM)-based multi-agent systems can automate alpha factor mining, but their reliance on external APIs limits control over cost, ava

local-aiarxiv-cs-ai

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…

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

agentsdair-ai--x

Beneath the Scores: Rethinking Hallucination Evaluation for Video Understanding Models

arXiv:2609.28991v1 Announce Type: cross Abstract: Video understanding is increasingly performed by multi-stage LLM agents that separate temporal grounding, visual observation, and reasoning. Yet these

agentsarxiv-cs-ai

Local Ai

Canopy: Exploiting Piecewise Smooth Tree Priors for Multi-Fidelity Bandits

arXiv:2609.30017v1 Announce Type: cross Abstract: Many LLM inference problems, including model routing, prefix-cache management, prompt trimming, and test-time search, can be viewed as optimization ov

local-aiarxiv-cs-ai

From Self-Distillation to Self-Practice: Privileged Information for Multi-Turn Agents

arXiv:2609.29051v1 Announce Type: new Abstract: On-policy self-distillation (OPSD) has become a popular recipe for post-training LLM agents. It supervises the agent model at the token level with a str

HiPACE: Hierarchical Phase-Boundary Analysis and Controlled Evaluation of Feature Absorption in Sparse Autoencoders

arXiv:2609.29551v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) decompose LLM activations into sparse dictionary atoms, so that each distinct concept gets its own feature. One recurring beh

IDRBench: Benchmarking the Interactive Capabilities of Deep Research Agents

arXiv:2601.06676v3 Announce Type: replace-cross Abstract: Large Language Model (LLM)-based deep research agents perform multi-step reasoning, web exploration, and long-form report generation. In these

IterSynth: Rethinking Deep Search Agents via Role-Decoupled Iterative Synthesis

arXiv:2609.29444v1 Announce Type: cross Abstract: Deep search requires LLM agents to decompose complex queries, search for evidence, and synthesize grounded answers, yet existing ReAct-style agents su

Large Language Models for Programming: Actually Fixing or Reimplementing Incorrect Code?

model-releasesarxiv-cs-cl

arXiv:2609.29410v1 Announce Type: new Abstract: Recent studies have shown that Large Language Models can effectively solve problems and fix bugs in diverse programming environments, including competit

Near-Oracle KV Selection via Pre-hoc Sparsity for Long-Context Inference

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

Spent the last two months seeing how far a non programmer (me) can get building a proper game with nothing but ChatGPT in a browser.

Gonna ironically preface this with: I haven't written the text below with an LLM. boxxy.io So I thought I'd show you what I've made over the last couple of months, because for me this has probably bee

model-releasesr-chatgpt

Unraveling the cognitive patterns of Large Language Models through module communities

arXiv:2508.18192v2 Announce Type: replace Abstract: Large Language Models (LLMs) have reshaped our world with significant advancements in science, engineering, and society through applications ranging

We announced our RSI Lab earlier this year: https://sakana.ai/rsi-lab/ Over the last two years, we have systematically shipped the foundatio…

We announced our RSI Lab earlier this year: https://sakana.ai/rsi-lab/ Over the last two years, we have systematically shipped the foundations for autonomous R&D: ▪ LLM²: AI automating research to inv

agentsdavid-ha--x