model releases

this talk by openai researchers going through hugging face incident is totally insane, so much to unpack openai only realized it was their a…

this talk by openai researchers going through hugging face incident is totally insane, so much to unpack openai only realized it was their agent who hacked hugging face infra while asking hf to revoke

model-releasesswyx--x

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Threshold-Based Early Stopping of Accumulations in Neural Networks with Binary Activation

arXiv:2608.06177v1 Announce Type: new Abstract: Binary neural networks are very attractive for constrained deployment, enabling small footprint and low-power inference. For binary activations, the dot

Toward Deployable Bangla Sign Language Recognition with Expert-Validated Data and a Lightweight Attention-Based Model

arXiv:2608.06252v1 Announce Type: cross Abstract: Deaf and hard-of-hearing people in Bangladesh communicate mainly through Bangla Sign Language (BdSL). Automatic BdSL recognition on personal devices c

TRAJDEBUG: Tracing Error Lifecycle to Identify Critical Failures in Long-Horizon Agent Trajectories

arXiv:2608.06346v1 Announce Type: new Abstract: LLM-based agentic systems have shown remarkable capabilities in complex domains, while suffering from cascading errors and difficulty in debugging. Crit

Tree-NET: Enhancing 2D Medical Image Segmentation Through Efficient Low-Level Feature Training

arXiv:2501.02140v2 Announce Type: replace-cross Abstract: This paper introduces Tree-NET, a novel framework for medical image segmentation that leverages bottleneck supervision to enhance both segment

TS-RAG: Retrieval Augmented Generation for Time Series Forecasting

arXiv:2608.06223v1 Announce Type: new Abstract: While deep learning models, particularly transformer-based architectures, have shown impressive performance in time series forecasting, the application

turns out you can get indirect prompt injection to ~0 on unseen attacks if you stack enough layers (model training + input probes + a classi…

model-releasesboris-cherny--x

turns out you can get indirect prompt injection to ~0 on unseen attacks if you stack enough layers (model training + input probes + a classifier checking intent). didn't expect that a year ago. auto m

Tytan: Interactive Neurosymbolic Construction of Analytic Semantic Schemas from Relational Data

arXiv:2608.06331v1 Announce Type: cross Abstract: From natural-language query interfaces to automated report generation, data analysis tools need a description of the data: the real-world entities it

Unified Agent: Managing Interactions across Devices

arXiv:2608.05729v1 Announce Type: new Abstract: As capabilities rapidly increase, AI agents can move from running inside one app to acting across a user's devices over time. Yet existing agent systems

Unifying Structured and Unstructured Data Insights with BQ Search Innovations

model-releasesgoogle-cloud-ai

Modern enterprises possess a vast amount of unstructured data, yet they frequently encounter significant challenges in managing and extracting value from it. Historically, unlocking the insights hidde

Universal Pathologies, Conditional Consequences: A Triple-Robustness Analysis of RAG for Multi-Hop Traceability

arXiv:2608.05153v1 Announce Type: cross Abstract: GraphRAG underperforms vector RAG on citation precision in many reports, but where and why have remained corpus-bound. We present a triple-robustness

upgraded my stack, and i can now work on almost anything from anywhere hands free: - talk to chief of staff (via remote codex voice or text)…

model-releasesyohei-nakajima--x

upgraded my stack, and i can now work on almost anything from anywhere hands free: - talk to chief of staff (via remote codex voice or text) - chief assigns tasks to managers of various projects - man

Vibe Compiler: A Research-Logic Synthesis Tool That Runs without Prompt Engineering -Toward Enhancing Metacognition for Sustaining Agency in the Age of Generative AI-

arXiv:2608.05545v1 Announce Type: cross Abstract: Generative AI used as a capable servant has greatly accelerated intellectual work, but it also risks eroding human epistemic agency by encouraging unc

VideoArgus: Agentic Rubric-Grounded Unified Evaluation for Video Generation and Editing

arXiv:2608.05485v1 Announce Type: new Abstract: Evaluating generated videos remains challenging because existing benchmarks rely on fixed evaluation content, cover only a subset of generation and edit

Visual Intention Grounding for Egocentric Assistants

arXiv:2504.13621v2 Announce Type: replace Abstract: Visual grounding associates textual descriptions with objects in an image. Conventional methods target third-person image inputs and named object qu

Vorch-Director: Interactive World Story Model via Noise-Aware Error Rectification

arXiv:2608.05776v1 Announce Type: new Abstract: Autoregressive continuation provides a natural path toward minute-scale audio-visual generation by repeatedly extending a short-window generator conditi

We analyzed DeepSeek-V4 Flash-0731 vs. GPT-5.6 Luna on software engineering tasks using DeepSWE. DeepSeek-V4 Flash-0731 delivers 80% of Luna…

model-releasestogether-ai--x

We analyzed DeepSeek-V4 Flash-0731 vs. GPT-5.6 Luna on software engineering tasks using DeepSWE. DeepSeek-V4 Flash-0731 delivers 80% of Luna’s performance at roughly 1/6 the cost. More insights in the

What Current AI Benchmarks Leave Unmeasured: Modality, Search, Citations, and Implications (for Safety Evaluations)

arXiv:2608.06202v1 Announce Type: cross Abstract: Large language model (LLM) benchmark evaluations are routinely used to support claims about model safety, reliability, and deployment readiness. Yet m

What Drives Test-Time Adaptation for CLIP? A Controlled Empirical Study from an Update Perspective

arXiv:2606.14299v2 Announce Type: replace Abstract: Vision-Language Models (VLMs) such as CLIP have become a standard backbone for open-vocabulary recognition, yet their zero-shot predictions remain v

When History Lies: Evaluating and Improving Tool Use under Misleading Multi-Turn Histories

arXiv:2608.06057v1 Announce Type: new Abstract: Tool-calling agents infer task state from accumulated dialogue and tool traces. In persistent interactions, however, historical traces may remain struct

When Self-Evolution Backfires: Pre-Commit Gating against Skill Contamination in LLM Agents

arXiv:2608.05810v1 Announce Type: new Abstract: Self-evolving agents accumulate capability by distilling reusable skills from their execution trajectories, but we find this process is not monotonic: p

Where Privacy Risk Lives in English-Source Multilingual RAG: A Stage-Decomposed Audit Across Five Query Languages

model-releasesarxiv-cs-cl

arXiv:2608.05163v1 Announce Type: new Abstract: A common assumption holds that switching to a non-English language makes a multilingual RAG system easier to attack for personal information. We test th

Who Checks the Citations? Benchmarking Legal Hallucination Detection

model-releasesarxiv-cs-cl

arXiv:2606.21155v2 Announce Type: replace Abstract: Attorneys, judges, and pro se filers increasingly use AI to draft legal documents, yet these tools frequently fabricate citations. Despite predictio

You can play it here: https://simonw.github.io/raccoon-heist-codex/ For comparison, here's Fable 5 + Claude Code's game, built from the exac…

model-releasessimon-willison--x

You can play it here: https://simonw.github.io/raccoon-heist-codex/ For comparison, here's Fable 5 + Claude Code's game, built from the exact same prompt https://x.com/simonw/status/208508951822360205

Zero-Shot Multi-Disease Labeling of Chest, Abdomen, and Pelvis CT Reports Using Open-Weight Large Language Models: The Effect of Labeling Conventions

model-releasesarxiv-cs-cl

arXiv:2506.03259v3 Announce Type: replace Abstract: Purpose: To compare five lightweight open-weight large language models (LLMs) with a rule-based algorithm (RBA) and fine-tuned RadBERT for zero-shot

2 x 5070ti Qwen 27B full config / stats

model-releasesr-localllama

Following up on yesterday's post about running everyone's faves on 2 x 16gb cards while maximizing performance and KV. Previous post data used abandoned Cu130 VLLM image. Stats here are done on cu129-

5.6 Sol much better in chat now and unlimited text chat for free users!

model-releasessam-altman--x

5.6 Sol much better in chat now and unlimited text chat for free users! We’re making better intelligence easier to access in ChatGPT for everyone: - GPT-5.6 Sol now powers both Instant and deep reason

A Counterexample to Fourier Alignment in Single-Neuron Modular Addition

arXiv:2608.04451v1 Announce Type: cross Abstract: We give a negative solution to MAIS-O60. We first construct an example in which an initially active ReLU neuron becomes completely inactive in finite

A Survey of Agent Memory in the Second Half: Towards Self-Evolving and Long-Horizon Agents

arXiv:2602.06052v4 Announce Type: replace-cross Abstract: Research in artificial intelligence is shifting from model innovations and benchmark scores towards problem definition and rigorous real-world

A Unified Model for Cross-Domain Clone Detection via Model Merging

arXiv:2608.04215v1 Announce Type: cross Abstract: The growing diversity of code clone types, from syntactic copies to cross-language semantic clones to AI-generated duplicates, has created a fragmenta

Above-ground Biomass Estimation with Geospatial Foundation Models

arXiv:2608.04792v1 Announce Type: new Abstract: Accurate estimation of Above-Ground Biomass (AGB) from satellite imagery is essential for the large-scale monitoring of carbon stocks, yet it remains a

ACA-GS: Adaptive-Capacity Anchored Gaussian Splatting for Compact Dynamic Radiance Fields

arXiv:2608.04581v1 Announce Type: new Abstract: Recent advances in 4D Gaussian Splatting (4DGS) enable high-fidelity, real-time spatiotemporal rendering, but expose a fundamental trade-off between mot

Active-SWE: Benchmarking Coding Agents for Proactive Bug Fixing without Issue Reports

arXiv:2608.04682v1 Announce Type: cross Abstract: Coding agents powered by large language models (LLMs) are increasingly adopted in software engineering (SWE) scenarios, capable of fixing a specific b

Adaptive Finite-Budget Training for CVaR Risk-Aware Q-Learning

arXiv:2608.04305v1 Announce Type: new Abstract: Risk-aware Q-learning (RaQL) provides a model-free, two-timescale estimator for dynamic risk objectives, but its finite-budget behavior remains fragile:

Advancing brain tumor research with privacy-first AI

model-releasesgoogle-cloud-ai

The intersection of medicine and AI has led to remarkable innovations. However, developers now face the thorny challenge of building robust medical AI tools that have been tested and evaluated on dive

Advancing Utility Pole and Sign Detection Through Deep Learning

arXiv:2608.04061v1 Announce Type: new Abstract: Utility poles are an essential part of the infrastructure used to support power distribution systems and other critical public services. Their regular i

Adversarial Attacks for Good: A Survey of Proactive Protection across the Visual Content Lifecycle

arXiv:2608.04314v1 Announce Type: cross Abstract: Once visual content enters an AI pipeline, its owner often retains little technical control over how it is used. Legal and regulatory remedies can add

Adversarially Robust Abductive Fusion of Pre-trained Transformer-based Perception Models

arXiv:2608.04190v1 Announce Type: new Abstract: Deploying pre-trained perception models in novel environments degrades their accuracy under distributional shift, and assembling them alone does not rec

Agentic Future Ready With BigQuery: Continually Improving Price-Performance, Zero Effort

model-releasesgoogle-cloud-ai

In the modern data landscape, query performance tuning and managing system price-performance is challenging, especially as the number of agentic workloads increase. Even for experienced developers and

An active-learning framework for real-time depth perception from monocular vision streams

arXiv:2608.04917v1 Announce Type: new Abstract: Biological visual systems can perceive depth from monocular vision flow, continuously integrating temporal visual cues while maintaining a balance betwe

An AI model from Meta also hacked another company during testing

model-releasessimon-willison

An AI model from Meta also hacked another company during testing Stop me if you've heard this one before: An AI model from the parent company of Facebook and Instagram hacked into another company’s sy

An Update to Sir Shortoken: Introducing LELP-S+ (Less English, Less Prose)

A small update to Sir Shortoken. Sir Shortoken already had Quick, Balanced, Deep, Bullets, and Aggressive Bullets. I wanted something between Bullets and normal prose. So I added LELP-S+ (Less English

model-releasesr-ollama

Anthropic will design its own hardware to power Claude

model-releasesars-technica

Anthropic is hiring a custom silicon team to design proprietary chips that will power its Claude models, while still planning a multi‑chip strategy that mixes internally designed hardware with compone

anyone confused about neurosymbolic AI—and its recent enormous victory—should read this. complete and total vindication for what I have been…

model-releasesgary-marcus--x

anyone confused about neurosymbolic AI—and its recent enormous victory—should read this. complete and total vindication for what I have been saying here all along. (see also my essays on Claude Code a

Anyone understand what the equivalent of GPT-5.6 Instant in ChatGPT is for the OpenAI API?

model-releasessimon-willison--x

The tweet is a question from user Simon Willison (posted on 6 Aug 2026) asking which OpenAI API model corresponds to the ChatGPT “GPT‑5.6 Instant” version. No answer or clarification is included in th

Argus: A General-Purpose Agentic Runtime for Long-Horizon Reasoning

arXiv:2608.05144v1 Announce Type: new Abstract: Long-horizon reasoning requires an agentic runtime that can persist when evidence supports its current approach and pivot when measurements reveal failu

ArtChart: Faithful Artistic Chart Generation with Integrated Text Rendering

arXiv:2607.16060v2 Announce Type: replace Abstract: Artistic charts combine data visualization with expressive marks, textures, and typography, but they are difficult for image generators: an output i

Attention-Only White-Box Transformer via LeJEPA-Based Self-Supervised Pretraining

arXiv:2608.04213v1 Announce Type: new Abstract: Existing studies on self-supervised learning for white-box networks typically decouple the derivation of white-box networks via optimization algorithms