model releases

Qwen next 3.8 and 3.8 27b Vs Sonnet 5.5 low and Sonnet 5.5 medium.

model-releasesr-localllama

Six months ago, a result like this was unthinkable. But now we can say it loud and clear: local models are at the cutting edge, and the gap of just a few months has been confirmed. Personally, I use Q

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

Qwen3-VL 8B on a laptop vs Opus 5.5 / Sonnet 5 / GPT-5.6 on 137 messy documents: beat GPT-5.6 on tax forms, lost badly on Indian date formats[R]

DGX agent

I benchmarked Qwen3-VL 8B Instruct (Q4_K_M, Ollama, M5 24GB, ~30s/doc) against Claude Opus 5.5, Sonnet 5 and GPT-5.6 Terra on: - receipts: CORD (Indonesia) and SROIE (Malaysia), 30 each - 20 scanned 1

model-releasesr-machinelearning

28 Sept 2026

RAZOR: Pruning Replaceable Experts in LLMs

model-releasesarxiv-cs-cl

arXiv:2609.30465v1 Announce Type: cross Abstract: Mixture-of-experts (MoE) models activate few experts per token but store the full expert pool. Expert pruning reduces this storage burden; at a fixed

Refining Cytology Predictions with Conditional Random Fields

arXiv:2609.31028v1 Announce Type: new Abstract: Vision-language models (VLMs) achieve strong zero-shot (ZS) classification on histology images but do not perform as well on cytology, whose stains and

RefRef: A Dataset and Benchmark for Reconstructing Refractive and Reflective Objects

arXiv:2505.05848v3 Announce Type: replace Abstract: Modern 3D reconstruction and novel view synthesis approaches have demonstrated strong performance on scenes with opaque, non-refractive objects. How

RoboMonitor: Label-Efficient Runtime Monitoring of Robot Task Execution via Predictive Representation Learning

model-releasesarxiv-cs-ro

arXiv:2609.30715v1 Announce Type: new Abstract: Learned robot policies produce actions, but their outputs alone do not establish whether execution is progressing as intended. Robot execution monitorin

Robust to Which Model Change? A Unified Evaluation of Robust Counterfactual Explanations

arXiv:2609.30918v1 Announce Type: cross Abstract: Robust counterfactual explanations promise recourse that still works after the model behind it changes. Whether they keep that promise depends on what

RupeeBias: Auditing Demographic Bias in Indian Economic Guidance from Large Language Models

model-releasesarxiv-cs-cl

arXiv:2609.31245v1 Announce Type: new Abstract: Individuals turn to large language models (LLMs) for guidance across a wide range of economic tasks, from comparing loan options and planning savings to

SAGE: A sampling-aware global evaluation benchmark for species distribution modeling

arXiv:2609.31082v1 Announce Type: new Abstract: Knowing where species occur is fundamental for biodiversity research and conservation. Species distribution models (SDMs) link species observations to e

SatNav: A Scalable Benchmark for Long-Horizon UAV Vision-Language Navigation from Satellite Imagery

arXiv:2609.31507v1 Announce Type: new Abstract: Urban uncrewed aerial vehicle (UAV) vision-language navigation (VLN) requires agents to follow instructions across extended urban spaces, inherently dem

Scaling Density Functional Theory with Gaussian Splatting

arXiv:2609.31483v1 Announce Type: cross Abstract: Density functional theory (DFT) strikes a practical balance between accuracy and computational cost in many problems of computational chemistry and ma

SciHorizon-eLab: An Agentic Protocol-to-Task Compiler for Scalable Benchmarking of Scientific Embodied Agents

arXiv:2609.30971v1 Announce Type: new Abstract: Embodied agents offer a promising route to automating scientific experimentation, yet their progress is constrained by the lack of reliable and systemat

SciR: A Controllable Benchmark for Scientific Reasoning in LLMs

arXiv:2606.13020v3 Announce Type: replace Abstract: Three paradigmatic forms of inference recur across scientific reasoning: deduction, induction, and causal abduction. Reliably evaluating LLMs on the

ScopeBench: Do Agents Preserve Engagement Boundaries Under Goal Pressure?

arXiv:2609.30325v1 Announce Type: new Abstract: Agents are increasingly deployed with real autonomy in web application and network penetration testing, where a single out-of-scope action can breach a

SeA-RVINS: Semantic-Aware Tightly Coupled RTK-Visual-Inertial System with Correlation-Preserving Robust Estimation for Urban Navigation

model-releasesarxiv-cs-ro

arXiv:2609.30814v1 Announce Type: new Abstract: Reliable absolute pose estimation in urban environments is undermined by outlier measurements and incorrect temporal associations that can persist in ti

Searching for 3.8 35B: Qwen3.6-35B-A3B (Testing 5 Finetunes vs. Base)

model-releasesr-localllama

TL;DR -- You should probably just use base Qwen3.6-35B, as only Occamy-1.0 is competitive with it. Tiel is a major let-down, worse than Ornith. KAT surprises (good), Nex surprises (bad). This post is

Seeing Semantic Shift: Difference-Aware Sentence-Level Temporal Segmentation of Sign Language Videos

arXiv:2609.31148v1 Announce Type: new Abstract: Recent advances in sign language understanding have achieved impressive success on short, single-sentence videos, yet their performance drops sharply wh

Self-Supervised Perceptually Interpretable Monocular Depth Estimation

arXiv:2609.30987v1 Announce Type: new Abstract: Self-supervised monocular depth estimation (MDE) enables depth prediction from monocular images without requiring ground-truth supervision, making it at

Self-Supervised Representation Learning: From Spectral Foundation Models to Auroral Emission Spectra

arXiv:2609.31206v1 Announce Type: new Abstract: Auroral spectrographs such as the Auroral Spectrograph In Skibotn (ASIS) record hundreds of thousands of emission spectra, but only a few hundred can be

Semantic Navigation for Issue Localization in Code Repository

arXiv:2609.31176v1 Announce Type: new Abstract: Repository-level issue localization aims to identify and rank the files and functions relevant to resolving a reported issue. LLM agents approach this t

SignTrace: Describe a Sign, Find the Word

arXiv:2609.30295v1 Announce Type: cross Abstract: Identifying an unfamiliar sign is difficult when a learner remembers its movement but does not know its meaning or formal feature codes. SignTrace add

SkillFlow: Scalable and Efficient Agent Skill Retrieval System

arXiv:2504.06188v3 Announce Type: replace Abstract: AI agents can extend their capabilities at inference time by loading reusable skills into context, yet equipping an agent with too many skills, part

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

Sonnet 5.5 is by far the best free model available right now

I know most people here probably don't care as much about sonnet 5.5 as they did opus 5.5, because their 100 and 200 plans already give them plenty of weekly usage for opus. But keep in mind: free use

model-releasesr-chatgpt

SPO: Discovering Adaptive Large Neighborhood Search Operators via Stackelberg Program Optimization

arXiv:2609.31179v1 Announce Type: new Abstract: Large neighborhood search (LNS) relies critically on destroy and repair operators, whose effectiveness depends on both adaptation to the evolving LNS st

Stale-Document Poisoning: When Outdated Retrieval Overrides Correct Model Answers

model-releasesarxiv-cs-cl

arXiv:2609.31342v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) is often used to address outdated knowledge by providing external evidence. But retrieval helps only when that evid

Starliner will get extra help from NASA, and if it can sort out its problems it'll fly with a crew in 2028, and also be adapted to fly on th…

model-releasesanthropic--x

Starliner will get extra help from NASA, and if it can sort out its problems it'll fly with a crew in 2028, and also be adapted to fly on the Vulcan. The uncrewed Starliner-1 mission is planned for th

Stealth Apart, Harm Together: Skill Cascading Attacks on Skill-Based Agent Systems

arXiv:2609.30383v1 Announce Type: new Abstract: A skill is a modular package of natural-language instructions, executable scripts, and reference resources that an agent can load at runtime to extend i

Storage-optimized Z4D machine family, now GA, is designed for IO-intensive workloads

model-releasesgoogle-cloud-ai

Today, we’re excited to announce the general availability of our next-generation Storage-optimized Z4D machine series in Google Compute Engine with both Virtual Machine (VM) and bare-metal instances.

STORM-Bench: Evaluating Online Video QA under Evolving and Incomplete Evidence

arXiv:2609.30981v1 Announce Type: new Abstract: Reliable online video question answering requires tracking state transitions while selectively abstaining when visual evidence is insufficient. Existing

Strategic Self-Consistency

arXiv:2609.30352v1 Announce Type: cross Abstract: Self-consistency has become a popular technique for enhancing the reasoning abilities of large language models by generating multiple reasoning paths

Structured Multitask Gaussian Processes for Probabilistic Full-Body Human Motion Prediction

model-releasesarxiv-cs-ro

arXiv:2603.07096v2 Announce Type: replace Abstract: Robots that share space with people need human motion predictions with reliable uncertainty, not just accurate means. This paper presents a structur

Structured Reasoning Agentic Framework for Interpretable Critical View of Safety Assessment

arXiv:2609.31524v1 Announce Type: new Abstract: Surgical scene understanding is critical for computer-assisted intervention, yet laparoscopic cholecystectomy remains challenged by the complex anatomy

Swift 1.5 + HyperQwen = 37% less task completion time at 100+ tps w/ 150k context on RTX 3090

model-releasesr-localllama

Hi everyone :) The amazing Swift finetunes of Qwen3.8 27B generate much fewer tokens at mostly similar benchmark performance to the original model, while repos like HyperQwen (formerly syv-ai/qwen38-2

Synth-JEPA: Joint Embedding Prediction for Renderer-Free Synthesizer Parameter Search

arXiv:2609.31024v1 Announce Type: cross Abstract: Sound matching can be formulated as optimizing synthesizer parameters against an audio-domain objective. However, objectives derived from generic audi

TemplateCraft: Agentic Visual Template Generation

arXiv:2609.31451v1 Announce Type: cross Abstract: The growing popularity of short videos has driven demand for one-click content creation. Visual templates turn uploaded images into personalized conte

The Hard Part Comes After Search: Benchmarking Web Agents on Synthesizing, Organizing, and Displaying Knowledge

arXiv:2609.30604v1 Announce Type: cross Abstract: Existing computer-use agent benchmarks do not fully evaluate agents acting as assistants. A useful assistant retrieves information across complex, mul

The KV Cache Is the New Memory Wall

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

The Linear Representation Hypothesis for Vision-Language-Action Models

arXiv:2609.30996v1 Announce Type: cross Abstract: The linear representation hypothesis (LRH) has become a standard lens for measuring and intervening on semantic information through the internal repre

The most important thing about Sonnet 5.5 is that it's now the model that powers the free tier on https://claude.ai - so all of this stuff c…

model-releasessimon-willison--x

The most important thing about Sonnet 5.5 is that it's now the model that powers the free tier on https://claude.ai - so all of this stuff can be done by free users ChatGPT's free tier is still GPT-5.

The Plot Twist: Jailbreaking Unified Multimodal Models with a Three-Act NarrativeAttack

arXiv:2509.26473v2 Announce Type: replace Abstract: Unified Multimodal Understanding and Generation Models (UMMs) increasingly combine visual understanding and image generation within a single interac

The Price of Thought: Does Test-Time Reasoning Pay in LLM Trading?

arXiv:2609.30705v1 Announce Type: new Abstract: While inference-time reasoning in large language models (LLMs) promises better decision making, its higher computational cost may not yield better econo

The Right Information Extraction Pipeline Depends on the Document: Accuracy-Energy Trade-offs for Small, Local Models

arXiv:2609.31341v1 Announce Type: new Abstract: Whether an information extraction pipeline should process page images or parsed text depends on the document, and the answer flips across the layout spe

The Shrinking Lifespan of LLMs in Science

arXiv:2604.07530v3 Announce Type: replace-cross Abstract: Scaling laws describe how language model capabilities grow with compute and data, but say nothing about how long a model matters once released

Tight Stochastic Condition-Number Dependence in Nonconvex-Strongly-Concave Minimax Optimization

arXiv:2609.30877v1 Announce Type: cross Abstract: We study whether the linear condition-number dependence in the stochastic complexity of SAPD+ is necessary for nonconvex-strongly-concave minimax opti

Timo: extbf{T}aming Multextbf{i}modal Diffusion Transformer for Human extbf{Mo}tion Generation

arXiv:2609.30761v1 Announce Type: new Abstract: Most existing human motion generation (HMG) methods use cross-attention modules to inject text semantics, but ignore the importance of bidirectional mod

Towards Understanding LLM-Based Log Anomaly Detection: An Empirical Study of Performance, Efficiency, and Robustness

arXiv:2609.31371v1 Announce Type: new Abstract: Large language models (LLMs) have demonstrated promising performance in log anomaly detection, yet how their adaptation strategies, architectures, and d

TR-SSQP: A Trust-Region Method for Constrained Stochastic Optimization under Heavy-Tailed Noise

arXiv:2609.30732v1 Announce Type: cross Abstract: We consider stochastic nonlinear optimization problems with deterministic equality constraints. While unconstrained stochastic optimization is well un