Search: “model-releases”

Your VLM Already Knows When: Training-Free Temporal Grounding by Asking Yes or No

arXiv:2608.08315v1 Announce Type: new Abstract: Multimodal LLMs that recognise events reliably still fail to say when they happen. Prompted for timestamps, strong VLMs reach as little as 3.8% R@0.5 on

11 Aug 2026

Model Releases

Zero-shot 2D Grounding with Novel Affordance Types

DGX agent

arXiv:2608.08929v1 Announce Type: new Abstract: 2D affordance grounding aims to locate the region of an object that a human can interact with. Existing research focuses on recognizing affordance types

model-releasesarxiv-cs-cv

11 Aug 2026

Zero-Shot Traffic Accident Detection via a Coarse-to-Fine VLM-Tracking Pipeline

arXiv:2608.08867v1 Announce Type: new Abstract: Traffic surveillance cameras capture accidents continuously, yet converting raw CCTV footage into structured event records that pinpoint when, where, an

11 Aug 2026

ZhuLong: Execution-Grounded LLM Agent for EDA Scripting with Offline API Self-Exploration

arXiv:2608.07925v1 Announce Type: new Abstract: EDA scripting with tool-specific, often undocumented APIs remains a long-tail bottleneck that existing LLMs fail to address. This paper presents ZhuLong

11 Aug 2026

1M context with 17 GB model in 24 GB VRAM: 'for the first time I was able to load a context of almost 1M tokens and extract 7 needles from various parts of the text'

https://preview.redd.it/xxjh11f38jih1.png?width=1852&format=png&auto=webp&s=76850ed51e29a8bc86c2ca718d4320075eed4363 Just wanted to share a user report that I found to be very interesting. Some person

8 days ago, while jogging, I asked Claude to solve the Riemann Hypothesis It didn’t. 1.5 days later, it proved >= 67% of the zeros are on th…

model-releasesboris-cherny--x

8 days ago, while jogging, I asked Claude to solve the Riemann Hypothesis It didn’t. 1.5 days later, it proved >= 67% of the zeros are on the line (prev: 41.6%) Still not sure what that means, but som

A downside with VLM-based parsing is that they’re generally slower than text-based heuristic approaches. As a result they add latency to any…

model-releasesjerry-liu--x

A downside with VLM-based parsing is that they’re generally slower than text-based heuristic approaches. As a result they add latency to any ad-hoc file processing *in-the agent loop* (e.g. if you upl

A Multi-Agent Framework for Automated Coarse-Grained Molecular Dynamics of Polymers

arXiv:2608.06694v1 Announce Type: new Abstract: Coarse-grained (CG) molecular dynamics extends polymer simulation beyond the scales accessible to all-atom (AA) methods, but bottom-up CG modeling is la

A Picture is Worth a Thousand Tokens: How Vision Language Models Cut AI Energy Costs While Improving Accuracy

arXiv:2608.07427v1 Announce Type: new Abstract: LLM inference accounts for over 90% of AI operational energy, scaling directly with input token count---a critical inefficiency for telecom network anal

Accounting Graph Transformer for Short-History Multi-KPI Forecasting in Small Businesses

arXiv:2608.07037v1 Announce Type: cross Abstract: Small businesses often have only 12-24 months of accounting history, yet planning and risk workflows require coordinated forecasts across financial st

Achievable 253 t/s - unsloth/Muse Glimmer 30B UD-Q5_K_M on a 5090

Benchmarked Muse Glimmer 30B on my RTX 5090 (32GB), 262k context, UD-Q5_K_M + dflash-kquant + mmproj. Workload Stock master + DFlash ngram-simple PR #26842 + DFlash Code patch 78 t/s 57 t/s 220-253 t/

Addressable Memory for Video World Models

arXiv:2608.07408v1 Announce Type: new Abstract: We study visual persistence in interactive video world models. These models rely on a Key-Value (KV) cache as a growing visual memory to carry forward p

Advanced capabilities require strong safeguards. That’s why access is limited to approved defenders, with additional controls and monitoring…

Advanced capabilities require strong safeguards. That’s why access is limited to approved defenders, with additional controls and monitoring for higher-risk cybersecurity work. https://openai.com/inde

model-releasesopenai--x

Aftab: A Comprehensive Benchmark of CNN Encoders and Advanced Value Functions in Parallelized Q-Networks

arXiv:2608.07335v1 Announce Type: cross Abstract: Recent advancements in deep reinforcement learning have increasingly favored simplified, highly parallelized paradigms. Notably, the Parallelized Q-Ne

Agent Memory Distillation: Empowering Small LLM Agents with Hierarchical Teacher Memory

arXiv:2608.07169v1 Announce Type: new Abstract: Memory systems have shown promise for improving agent performance, but their potential remains largely unexplored for small language models, which strug

An Exploratory Evaluation of LLM-Assisted Rewriting of Moderate-Complexity Financial Sentences for DisCoCat-Based Sentiment Analysis

model-releasesarxiv-cs-cl

arXiv:2608.07439v1 Announce Type: new Abstract: Quantum natural language processing (QNLP) provides a grammar-aware framework for text modeling, and Distributional Compositional Categorical (DisCoCat)

Anthropic details an unreleased Claude model's attempt to solve the Riemann hypothesis; it didn't solve it but 'unexpectedly' made strides on a related problem (Anthropic)

Anthropic: Anthropic details an unreleased Claude model's attempt to solve the Riemann hypothesis; it didn't solve it but “unexpectedly” made strides on a related problem — Recently, a member of staff

model-releasestechmeme

Anthropic, Macquarie, and Singapore's GIC form Theseus Infrastructure to develop AI computing sites; Anthropic commits to cover consumer electricity price hikes (Mark Bergen/Bloomberg)

Mark Bergen / Bloomberg: Anthropic, Macquarie, and Singapore's GIC form Theseus Infrastructure to develop AI computing sites; Anthropic commits to cover consumer electricity price hikes — Anthropic PB

model-releasestechmeme

Anthropic makes Claude Sonnet 5's introductory pricing permanent at 2/1M input tokens and 10/1M output tokens, canceling a planned September 1 increase (Claude/@claudeai)

Claude / @claudeai: Anthropic makes Claude Sonnet 5's introductory pricing permanent at 2/1M input tokens and 10/1M output tokens, canceling a planned September 1 increase — We're making Claude Sonnet

model-releasestechmeme

ArchEGraph: A Large-Scale Graph Dataset for Geometry-Topology-Physics Aligned Building Energy Modeling

arXiv:2608.06772v1 Announce Type: new Abstract: Accurate estimation of building energy use is essential for achieving carbon neutral and sustainable buildings. To better understand the influence of de

Artificial Intelligence Can Match Domain Experts in Evidence Extraction and Critical Appraisal of Microbial Oncogenesis Research Publications

arXiv:2608.07250v1 Announce Type: cross Abstract: Confirmed oncogenic microbes contribute significantly to cancer burden. Identifying novel microbial oncogenicity could yield strategies that will redu

Ask-E: An Environment for Calibrated Question Generation

arXiv:2608.06933v1 Announce Type: cross Abstract: Today, we improve models by training and evaluating them on problems at the frontier of their abilities. Creating such problems is itself a demanding

AutoMOOSE: An Agentic AI for Autonomous Phase-Field Simulation

arXiv:2603.20986v2 Announce Type: replace Abstract: Phase-field modeling links thermodynamics and kinetics to microstructural evolution, but multiphysics frameworks such as MOOSE require expertise to

AVCap: Reinforcing Audio-Video Joint Caption with Detail-Aware Reward

arXiv:2608.06930v1 Announce Type: new Abstract: Detailed audio-video joint captioning is essential for multimodal video understanding and generation. However, prior works are constrained by three main

b10336

ggml-webgpu : refactor several wgsl files and simplify flash_attn wgsl. (#26134) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLE

b10338

model-saver : fix expert shared/chunk FFN length key clobber (#26693) The saver called add_kv with LLM_KV_EXPERT_SHARED_FEED_FORWARD_LENGTH twice, the second time passing n_ff_chexp. gguf_set_val_u32

b10342

model : Granite-Switch Architecture (#25107) granite-switch: add llama.cpp backend (POC, CPU) New 'granite-switch' architecture: a dense, all-attention Granite-4.1 model with N embedded LoRA adapters

b10343

vendor : update cpp-httplib to 0.53.0 (#26821) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFramewor

b10344

model: add MTP support for Nemotron model (#26725) model: add MTP support for Nemotron Nano model model: add mtp_flags for nemotron model address review comments Website: https://llama.app macOS/iOS:

b10353

ggml : require contiguous src for ROLL on CUDA and Metal (#25928) ggml_roll only asserts nb[0] == ggml_type_size, so a permuted src is a valid input, but the CUDA and Metal roll kernels index by ne al

b10354

ggml-cpu : fix CPU affinity mask being ignored on Android (#26838) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel

b10355

llama : support multi-output backend sampling (#25532) Enable backend sampling with token speculation Clamp the mask sum before converting it into the sampled index Add a numeric context parameter dec

Benchmarking and Reasoning Distillation of Large Language Models for Feedback Controller Design in Complex Dynamical Systems

model-releasesarxiv-cs-ro

arXiv:2608.07004v1 Announce Type: new Abstract: Although remarkable capabilities have been demonstrated by Large Language Models (LLMs) across scientific domains, feedback controller design remains un

Best open-source harness like Claude Code?

Avid claude code user here looking to do equivalent things with local models. Just want to plug in something like Qwen and have the interface be 1:1 with claude code. Any suggestion? submitted by /u/N

Beyond Fluency: A Clinical Benchmark and Anomaly-Enhanced Baseline for Spine MRI Report Generation

arXiv:2608.07117v1 Announce Type: new Abstract: Radiology reporting is time-consuming and subject to inter-rater variability, making automated report generation an attractive clinical application for

Beyond Foundation Models: Dimension-Aware Neural Architecture Search with Small-Data Representation Models for Cryocooler Lifetime Prediction

arXiv:2608.06993v1 Announce Type: cross Abstract: Large-scale pretrained time-series models achieve strong results through large-scale pretraining and task-agnostic representation learning, but they r

Beyond Starry Night: Shortcut-Aware Control-State Planning for Artist-Grounded Text to Image Generation

arXiv:2608.06751v1 Announce Type: cross Abstract: Artist-grounded image generation requires more than appending an artist name to a prompt. Image models often respond to artist names through canonical

Beyond Text Matching: Towards Reference-Free Evaluation for Human-Oriented Binary Reverse Engineering

arXiv:2608.07038v1 Announce Type: cross Abstract: Human-Oriented Binary Reverse Engineering (HOBRE) aims to transform decompiled pseudocode into a more human-friendly representation, thereby reducing

Beyond the Black Box: Interpretable Models of Human Randomisation Failures

arXiv:2608.07220v1 Announce Type: new Abstract: Mixed strategy equilibrium predicts i.i.d play: past actions should not help predict future decisions. Human players, however, systematically depart fro

CADSpotting: Robust Panoptic Symbol Spotting on Large-Scale CAD Drawings

arXiv:2412.07377v5 Announce Type: replace Abstract: We introduce CADSpotting, an effective method for panoptic symbol spotting in large-scale architectural CAD drawings. Existing approaches often stru

Can Language Models Imagine Without Seeing? Ekphrasis: Measuring Visual Creative Ideation in Text-Only LLMs

model-releasesarxiv-cs-cl

arXiv:2608.06967v1 Announce Type: new Abstract: Current evaluations do not isolate whether text-only language models can originate visual concepts before image generation. Fluent visual prose can hide

Capacity Confounds and Coverage Guarantees in Adaptive Sub-model Federated Learning

arXiv:2608.07157v1 Announce Type: new Abstract: Sub-model federated learning lets resource-constrained clients train width-reduced versions of a global model, but existing methods allocate capacity by

Capek 0.5: An Execution-Centric Vision-Language Model for Embodied Intelligence

arXiv:2608.06756v1 Announce Type: new Abstract: Vision-language models are increasingly serving as the reasoning core of embodied agents. Robot execution is inherently iterative: each action reshapes

Cascading Through the Hierarchy: Regularizer-Induced Feature Detection as Phase Transitions in Deep Linear Neural Networks

arXiv:2608.06597v1 Announce Type: cross Abstract: A scientific theory of deep learning, comprising learning dynamics and statistical properties of learned models, is rapidly gaining attention. One of

Certified Interpolation Oversampling: Per-Instance Safety Guarantees for Imbalanced Learning

arXiv:2501.15790v2 Announce Type: replace Abstract: Synthetic minority oversampling is typically designed and evaluated against a predictive objective, generating samples that improve downstream class

Chat UIs with native audio input for multimodal models?

I've been running Gemma 4 E4B with oMLX and I can't find any chat interfaces that directly send the audio file to the model instead of running the audio through a separate STT layer. I can confirm the

Claude Haiku is my current least favorite model - it hallucinates wildly, and is out-performed now by other similarly priced models like GPT…

model-releasessimon-willison--x

Claude Haiku is my current least favorite model - it hallucinates wildly, and is out-performed now by other similarly priced models like GPT-5.6-Luna Even worse: it seems to still be used by the Claud

Cluster Attention for Graph Machine Learning

arXiv:2604.07492v2 Announce Type: replace-cross Abstract: Message Passing Neural Networks have recently become the most popular approach to graph machine learning tasks; however, their receptive field