model releases

FreqAdapt: Frequency-Adaptive Processing for RAW Object Detection

arXiv:2608.03385v1 Announce Type: new Abstract: Existing object detection methods predominantly utilize sRGB inputs, which are compressed from RAW sensor data using Image Signal Processors (ISP) origi

Model Releases

From Generator to Embedder: Harnessing Innate Abilities of Multimodal LLMs via Building Zero-Shot Discriminative Embedding Model

DGX agent

arXiv:2508.00955v3 Announce Type: replace-cross Abstract: Adapting generative Multimodal Large Language Models (MLLMs) into universal embedding models typically demands resource-intensive contrastive

model-releasesarxiv-cs-ai

5 Aug 2026

From Social Coding to Agentic Coding: Productivity and Relational Reconfiguration in Open-Source Communities

arXiv:2608.03585v1 Announce Type: new Abstract: Open-source software communities are a form of digital public infrastructure that not only produces code, but also generates public knowledge and interp

Frozen High-Resolution Inference for Cross-City Object Detection: An AI City Challenge 2026 Study

arXiv:2608.03136v1 Announce Type: new Abstract: Cross-city object detection requires a detector trained in one city to generalize to an unlabeled target city. In AI City Challenge 2026 Track 6, we ana

Fusion-Poly: A Polyhedral Framework Based on Spatial-Temporal Fusion for 3D Multi-Object Tracking

arXiv:2603.08199v2 Announce Type: replace Abstract: LiDAR-camera 3D multi-object tracking (MOT) combines rich visual semantics with accurate depth cues to improve trajectory consistency and tracking r

GDPevo: Evaluating Agent Self-Evolution on Real Business Tasks

arXiv:2608.03764v1 Announce Type: new Abstract: Agent self-evolution updates an agent's persistent state from prior experience and reuses it to solve related tasks more effectively. Evaluating self-ev

GENESIS: Towards Explainable Causal Discovery

arXiv:2608.03868v1 Announce Type: cross Abstract: Causal Discovery (CD) from observational data faces two fundamental challenges. First, purely statistical methods often lack the power to resolve stru

Geo-Embed: Towards Unified Multimodal Embeddings for Urban Understanding

arXiv:2608.03826v1 Announce Type: new Abstract: Geospatial and urban applications increasingly require models to compare heterogeneous evidence across street-view imagery, remote-sensing observations,

Getting the Parameters Right: A Difficulty-Graded Benchmark and Probe-Guided Training for LLM Tool Calls

arXiv:2608.03071v1 Announce Type: new Abstract: Large language model agents derive much of their capability from tool use. Existing research on tool use has largely focused on selecting the right tool

GoT-CD: Graph-of-Thoughts Causal Discovery and the Fragility of Post-hoc Path-Specific Fairness Audits

model-releasesarxiv-cs-lg

arXiv:2608.02877v1 Announce Type: new Abstract: Causal discovery recovers directed structure from observational data and is increasingly used in clinical settings to support mechanism reasoning and fa

GUI-Lens: Coarse-to-Fine Cropping for GUI Grounding with General-Purpose VLMs

arXiv:2608.03270v1 Announce Type: cross Abstract: GUI grounding maps natural-language instructions to click locations and is essential for reliable GUI agents. The task remains difficult on high-resol

Hark, founded by Figure AI CEO Brett Adcock, previews Handoff, a computer use agent it says outperforms GPT-5.4 and Opus 4.8, and plans for a summer release (Ivan Mehta/TechCrunch)

Ivan Mehta / TechCrunch: Hark, founded by Figure AI CEO Brett Adcock, previews Handoff, a computer use agent it says outperforms GPT-5.4 and Opus 4.8, and plans for a summer release — Hark, a startup

model-releasestechmeme

Harness choice is a big deal. So much room to advance and improve results across the board with agent harnesses. Great paper highlighting th…

Harness choice is a big deal. So much room to advance and improve results across the board with agent harnesses. Great paper highlighting this. New research releases DataSpace, a benchmark where data

model-releasesdair-ai--x

Hey everyone! I just released a sub-6B sparse activation AI model which was built with a brand new architecture : fusion. I fused weights fr…

model-releasesemad-mostaque--x

Hey everyone! I just released a sub-6B sparse activation AI model which was built with a brand new architecture : fusion. I fused weights from @liquidai's LFM2.5-2.6B & @Alibaba_Qwen's Qwen3.6-35B-A3B

HomeSafeBench: A Benchmark for Embodied Vision-Language Models in Free-Exploration Home Safety Inspection

arXiv:2509.23690v2 Announce Type: replace-cross Abstract: Safety hazards in the home are a leading cause of preventable domestic injuries, motivating an automated inspector that actively explores a ho

How Closely Do LLM Reviews Align with Human Peer Review?

arXiv:2608.03659v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used to generate scientific reviews, yet existing evaluations rarely examine whether different providers

HUKUKBERT: Domain-Specific Language Model for Turkish Law

arXiv:2604.04790v2 Announce Type: replace Abstract: Natural language processing (NLP) advances have powered a generation of LegalTech systems, but Turkish law remains under-served by domain-specific d

HyperFL: Query-Adaptive Representation Learning for Software Fault Localization

arXiv:2608.02967v1 Announce Type: cross Abstract: Software fault localization identifies the code locations responsible for reported issues and is a fundamental step toward automated debugging and pro

HyVIC: A Metric-Driven Spatio-Spectral Hyperspectral Image Compression Architecture Based on Variational Autoencoders

arXiv:2603.26468v2 Announce Type: replace Abstract: The rapid growth of hyperspectral data archives in remote sensing (RS) necessitates effective compression methods for storage and transmission. Rece

I remember a time when 'flash' meant 32B

I mean, Deepseek V4 Flash is an absolutely fantastic model, even though I can't run it on my machine it's so fascinating to see how it performs. Knowing that potentially it could be run at home is rea

I updated my localy run benchmark with DeepSeek V4 Flash 0731

It's the purple cluster on the top left (the good corner...) I'm running the MXFP4 version from Bartoswski with Dspark at 1K t/s prefill and 90 t/s gen (average). I tried different sampling params, yo

In-Context Collapse in Vision-Language Models and How to Mitigate it?

arXiv:2608.02830v1 Announce Type: cross Abstract: Many-shot in-context learning (ICL) lets vision-language models (VLMs) adapt from image--label demonstrations without weight updates, and is widely as

In-Context Pure Exploration in Continuous Decision Spaces

arXiv:2602.17976v2 Announce Type: replace-cross Abstract: In active sequential testing, also termed pure exploration, a learner is tasked with the goal to adaptively acquire information so as to ident

Incident Report: unsanctioned agent behaviour during cyber testing

model-releasessimon-willison

Incident Report: unsanctioned agent behaviour during cyber testing It happened again. This time it was the UK government's AI Security Institute who accidentally attacked other companies while running

Inkling-Small 276B-A12B at ~2.9 tok/s on <10gb memory

A follow up to the launch of Mference, it now supports and runs Inkling-Small 276B-A12B. Inkling-Small (Thinking Machines, Apache 2.0), from the pipenetwork/Inkling-Small-MLX-4bit conversion: 276B tot

Instruction Stacking Collapse: A Benchmark and the Capability-Dependent Value of Prompt Compilation

arXiv:2608.02639v1 Announce Type: cross Abstract: Production prompts rarely carry a single instruction. One system message may require valid JSON, a word limit, three citations, and a fixed tone at th

Internalizing Academic Writing Workflows for Introduction Generation via Struct-Aware Policy Learning

arXiv:2608.03138v1 Announce Type: cross Abstract: Generating a rigorous paper introduction with large language models (LLMs) remains challenging, since it requires coordinating background, gap identif

Intertemporal Preference Steering in Qwen3 via Contrastive Activation Addition

arXiv:2608.03892v1 Announce Type: new Abstract: We study linear representations of temporal horizon in the large language model Qwen3-32B and use them to change the model's time-related preferences, r

Introducing Muse Code and Muse Spark 1.2

model-releasessimon-willison

Introducing Muse Code and Muse Spark 1.2 Yet more evidence that the most important characteristic of any model these days is long-sequence agentic tool calling. Meta shipped their own coding agent as

Inverted Detection and Control in Steering Vectors

model-releasesarxiv-cs-lg

arXiv:2608.02957v1 Announce Type: new Abstract: Steering vectors (SVs) are widely used to influence the expression of concepts (e.g., truthfulness) in large language model outputs. A key assumption un

IPPRO: Importance-based Pruning with PRojective Offset for Magnitude-indifferent Structural Pruning

arXiv:2507.14171v3 Announce Type: replace-cross Abstract: Importance-based structured pruning overwhelmingly relies on filter magnitude. This proxy is fundamentally flawed: due to scale invariance, fu

JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion

arXiv:2608.03974v1 Announce Type: new Abstract: Real-time video editing requires low-latency causal generation with bounded computational resources while preserving source fidelity and long-term tempo

JudgeArena: A Unified Framework for Reproducible LLM-Judge Evaluation

arXiv:2608.02620v1 Announce Type: new Abstract: LLM-as-a-judge evaluation has become a dominant paradigm for ranking language models, yet the ecosystem remains fragmented: most benchmarks ship their o

KnowHal: A Knowledge-Driven Benchmark for Comprehensive Multimodal Hallucination Evaluation

arXiv:2608.03782v1 Announce Type: new Abstract: Hallucination remains a critical challenge for developing trustworthy Multimodal Large Language Models (MLLMs). While existing benchmarks mainly focus o

Knowing the Form, Not the Function: Automatically Auditing Answer--Authority Decoupling in Legal Benchmarks

arXiv:2608.02621v1 Announce Type: cross Abstract: Legal benchmarks typically score final answers even when models also state legal authority. We test whether answer correctness can serve as a proxy fo

LAEF: A Lead-Agnostic ECG Foundation Model Towards Point-of-Care Diagnostics

model-releasesarxiv-cs-lg

arXiv:2608.03690v1 Announce Type: new Abstract: Point-of-care cardiac devices such as smartwatches and handheld ECG recorders typically capture 1--2 leads, yet existing ECG foundation models are archi

Latent Reward Registers for Diffusion Preference Alignment

arXiv:2608.03929v1 Announce Type: cross Abstract: Aligning diffusion models with human preferences usually relies on a sparse terminal reward evaluated on the final generated samples, presenting a sev

LDU-Bench: Multimodal LLM Evaluation for Lithography Defect Understanding under Layout-Varying Circuit Backgrounds

arXiv:2608.03078v1 Announce Type: new Abstract: Multimodal large language models have demonstrated strong defect recognition capability in industrial anomaly detection. However, in lithography review,

LeanMem: Simple and Efficient Long-Term Memory for LLM Agents

arXiv:2608.03463v1 Announce Type: new Abstract: Long-term memory is essential for LLM-based agents to sustain interactions and reliably leverage distant history. However, existing memory systems typic

Let's create with Qwen-Image-3.0-Pro on fal! 🎨

Let's create with Qwen-Image-3.0-Pro on fal! 🎨 Qwen Image 3.0 Pro is now available on fal - Renders complex, dense typography - Preserves key details like facial features and identity while applying c

model-releasesqwen--x

LFM2.5-2.6B on a OnePlus 13 at 17 tok/s ~ Pure CPU

As you all know the model is 2.69B parameters with a 128K context window and purpose-built for multi-step agent workflows. What you are seeing is the Q4_K_M GGUF running on my own inference engine bui

Lightweight Chunk Selection for Mobile Retrieval-Augmented Generation

arXiv:2608.03148v1 Announce Type: cross Abstract: RAG improves the factual grounding of LLM by incorporating external knowledge, but deploying RAG on mobile and edge devices remains challenging becaus

LoCA: Forward-Only LLM Tuning after One-Shot Calibration with Local Credit Assignment

arXiv:2608.03020v1 Announce Type: new Abstract: Parameter-efficient post-training reduces the number of trainable parameters, but still requires repeated end-to-end backpropagation through the frozen

Localize, Don't Beautify: Client-Side Control of Image-Editing APIs for Cosmetic Surgery Previews

arXiv:2608.02841v1 Announce Type: new Abstract: Ask a commercial image editor to preview a cosmetic procedure and it will often change more of the face than the request names: a nose edit can also smo

LocAnyMed: Vision-Language Grounding for Multimodal Medical Images

arXiv:2608.03322v1 Announce Type: new Abstract: Medical visual grounding connects free-form clinical queries to spatial evidence in medical images and is an important component of interpretable medica

Looking under the Wrong Lamppost: On the Limitations of Automated Translation Quality Estimation

arXiv:2608.03577v1 Announce Type: new Abstract: Automation of Translation Quality Estimation (QE) has emerged as a widely discussed approach to managing translation quality at scale, and a growing num

LoopMTP: A looped transformer guided by latent multi-token prediction

arXiv:2608.03624v1 Announce Type: new Abstract: Looped transformers have emerged as a parameter-efficient alternative to scaling depth for strong reasoning. By reusing one stack of layers across T ite

Low-Dimensional High-Leverage Subspace Optimization: Beyond Full-Parameter Coupled Training for Neural Network Quantization

arXiv:2608.03919v1 Announce Type: new Abstract: Low-bit quantization suffers severe accuracy degradation on compact networks, rooted in the dominant full-parameter coupled training paradigm that ignor