All entries

The creation of a regulatory body to enforce and implement standards, as suggested here by @demishassabis or by @AnthropicAI last month, wou…

The creation of a regulatory body to enforce and implement standards, as suggested here by @demishassabis or by @AnthropicAI last month, would be a huge step in the right direction to make AI developm

safetyyoshua-bengio--x

Agents

The Emerging Paradigm of Geospatial Foundation Models: From Pre-Training to Agentic Reasoning

DGX agent

arXiv:2607.12177v1 Announce Type: new Abstract: The analysis of satellite and aerial imagery has entered a new era with the advent of foundation models. This paper describes the concept of Geospatial

agentsarxiv-cs-ai

15 Jul 2026

The Geometry of Memorization: Finite-Time Spectral Sensitivity as a Diagnostic for Flow Matching Models

arXiv:2607.12616v1 Announce Type: new Abstract: Continuous-time generative frameworks construct probability paths between base and target domains by optimizing time-dependent velocity fields. While th

researcharxiv-cs-lg

The GEST-Engine: From Event Graphs to Synthetic Video. A Full Technical Report

arXiv:2607.12231v1 Announce Type: new Abstract: We present the GEST-Engine, a complete system that goes from natural-language text to fully-annotated multi-actor video. At its core is an explicit worl

agentsarxiv-cs-cv

The Illusion of Robustness: Aggregate Accuracy Hides Prediction Flips under Task-Irrelevant Context

model-releasesarxiv-cs-cl

arXiv:2607.12963v1 Announce Type: new Abstract: As large language models (LLMs) grow more capable, they are increasingly deployed in context-rich settings where task inputs are often accompanied by lo

The Limits and Potentials of Local SGD for Distributed Heterogeneous Learning with Intermittent Communication

arXiv:2405.11667v2 Announce Type: replace Abstract: Local SGD is a popular optimization method in distributed learning, often outperforming other algorithms in practice, including mini-batch SGD. Desp

researcharxiv-cs-lg

The Model Knows Your Project, Not You: Measuring Recognition in LLMs with NameRank

arXiv:2607.12520v1 Announce Type: new Abstract: What a frontier model recalls about a person or tool from its own weights -- before any retrieval step -- often shapes the first description a human see

researcharxiv-cs-ai

The One-Word Census: Answer-Choice Conformity Across 44 Language Models

arXiv:2607.12796v1 Announce Type: cross Abstract: When a language model must pick one answer from a large space of equally valid options, which does it pick -- and how often is it the same answer ever

The Risk of Exposed Cloud Functions and How to Harden

model-releasesgoogle-cloud-ai

Written by: Corné de Jong Introduction Mandiant security assessments frequently identify publicly exposed serverless applications that lack authentication, often as a result of specific business requi

The Seriality Gap in Video Diffusion Models

arXiv:2607.13031v1 Announce Type: cross Abstract: When one ball strikes another, then another, video models should predict the consequences of each bounce. In controlled experiments on multi-ball hard

researcharxiv-cs-cv

The Sound of Absence: Audio-Language Embedding Models Struggle with Negation

arXiv:2607.12290v1 Announce Type: cross Abstract: Audio-language embedding models such as CLAP are widely evaluated on matching present sound events, but rarely on negation. We show this affirmation-o

safetyarxiv-cs-ai

The Spectrum Is Not Enough: When Context Helps Time-Series Forecasting

arXiv:2607.13006v1 Announce Type: new Abstract: A growing family of indices scores how predictable a series is from its spectrum. Practitioners increasingly read these scores as answering a different

researcharxiv-cs-lg

The TopCoW Challenge -- Topology-Aware Circle of Willis Segmentation for CT and MR Angiography

model-releasesarxiv-cs-cv

arXiv:2312.17670v5 Announce Type: replace Abstract: The Circle of Willis (CoW) is an important network of arteries connecting major circulations of the brain. Its vascular architecture is believed to

Thinking Machines just dropped a ~1T Omni model 🔥 > 1M context window, trained on 48T tokens of image, text, audio 🤯 > comes with a drafte…

model-releasesclem-delangue--x

Thinking Machines just dropped a ~1T Omni model 🔥 > 1M context window, trained on 48T tokens of image, text, audio 🤯 > comes with a drafter, NVFP4 weights and Unsloth quants > transformers, llama.cpp

Industry

Thinking Machines Lab debuts Inkling, an open-weight MoE model with 975B total and 41B active parameters, trained to be broad rather than optimized for one area (Thinking Machines Lab)

Thinking Machines Lab: Thinking Machines Lab debuts Inkling, an open-weight MoE model with 975B total and 41B active parameters, trained to be broad rather than optimized for one area — Try on Tinker

industrytechmeme

Thinky with a ~1T param, 41B active, apache-2 model Benchmarks are a clear step up from Nemotron Ultra (55B active), new best American model…

model-releasesclem-delangue--x

Thinky with a ~1T param, 41B active, apache-2 model Benchmarks are a clear step up from Nemotron Ultra (55B active), new best American model, and omni input. A bit behind GLM 5.2 on agentic benchies,

This AI recursive self improvement (RSI) paper shows no sign of fast takeoff. The AI model advances at about [Intelligence]^0.075, or the th…

This AI recursive self improvement (RSI) paper shows no sign of fast takeoff. The AI model advances at about [Intelligence]^0.075, or the the 13th root of input intelligence [1,2]. That means the inte

safetygary-marcus--x

This is for pretraining based on nemotron nvp4 figures, mfu etc Large context, multimodal, long RL we are seeing now could make it a multipl…

model-releasesemad-mostaque--x

This is for pretraining based on nemotron nvp4 figures, mfu etc Large context, multimodal, long RL we are seeing now could make it a multiple of this Worth noting how good the lite model is, similar t

This is insane! We've just started the challenge, and already 73 papers from ICML are partially or fully reproduced thanks to the community!…

This is insane! We've just started the challenge, and already 73 papers from ICML are partially or fully reproduced thanks to the community! Join the effort here 👉 https://huggingface.co/spaces/ICML-2

agentsclem-delangue--x

Hardware

This piece by @GavinSBaker makes me wonder if @elonmusk might consider making future @grok models 'open.' 1- Like @nvidia they have the reso…

This piece by @GavinSBaker makes me wonder if @elonmusk might consider making future @grok models 'open.' 1- Like @nvidia they have the resources to maintain an open model. 2- They have activity in ma

hardwareclem-delangue--x

this was such a fun talk with Cat and Simon, hope you get to check it out if you missed it

this was such a fun talk with Cat and Simon, hope you get to check it out if you missed it 🆕This Year In Claude https://www.youtube.com/watch?v=uU5Gv2h8-9g @simonw chats with @_catwu and @trq212 about

model-releasesthariq--x

🆕This Year In Claude https://www.youtube.com/watch?v=uU5Gv2h8-9g @simonw chats with @_catwu and @trq212 about the state of: - @claudeai Cod…

🆕This Year In Claude https://www.youtube.com/watch?v=uU5Gv2h8-9g @simonw chats with @_catwu and @trq212 about the state of: - @claudeai Code - Claude Fable - @anthropicai culture & product strategy -

model-releasesswyx--x

Thompson Sampling Is 2-Competitive for Mistakes

arXiv:2607.12389v1 Announce Type: cross Abstract: We consider Bayesian bandit models and prove that Thompson sampling makes at most twice the expected number of mistakes (selections of a suboptimal ar

safetyarxiv-cs-lg

Applications

Tilt-Ropter: A Fully Actuated Hybrid Aerial-Terrestrial Vehicle with Tilt Rotors and Passive Wheels

arXiv:2602.01700v3 Announce Type: replace Abstract: In this work, we present Tilt-Ropter, a fully actuated hybrid aerial-terrestrial vehicle (HATV) that integrates tilt rotors with passive wheels to e

applicationsarxiv-cs-ro

Industry

Today, we are introducing Inkling. Inkling reasons efficiently across text, image, and audio modalities. We are making the full weights avai…

Today, we are introducing Inkling. Inkling reasons efficiently across text, image, and audio modalities. We are making the full weights available. https://thinkingmachines.ai/news/introducing-inkling/

industryclem-delangue--x

Together, Then Apart: Balancing Alignment and Distinctiveness for Multimodal Survival Analysis

arXiv:2511.18089v2 Announce Type: replace Abstract: Multimodal survival analysis aims to improve cancer prognosis using heterogeneous biomedical data, such as histopathology images and genomic profile

safetyarxiv-cs-cv

Token Reduction Is Not Cost Reduction

model-releasesarxiv-cs-cl

arXiv:2607.12161v1 Announce Type: new Abstract: Context-reduction layers for API-based coding agents, including command-output compressors, retrieval rankers, and payload-optimizing proxies, are usual

Local Ai

Toward Localizing and Repairing Bias in Transformer Attention Heads

arXiv:2607.12863v1 Announce Type: cross Abstract: Transformer language models are increasingly used as software components, yet biased outputs remain difficult to localize and repair inside the model.

local-aiarxiv-cs-lg

Toward Metaphor-Fluid Conversation Design for Voice User Interfaces

arXiv:2502.11554v3 Announce Type: replace-cross Abstract: Metaphors play a critical role in shaping user experiences with Voice User Interfaces (VUIs), yet existing designs often rely on static, human

researcharxiv-cs-ai

Local Ai

Toward Production-Ready Federated Learning in Healthcare: Privacy, Orchestration, and Governance in MLOps

arXiv:2607.10467v2 Announce Type: replace-cross Abstract: Healthcare organizations often cannot freely centralize patient data because medical records are sensitive, regulated, and institutionally con

local-aiarxiv-cs-lg

Toward Trustworthy Autonomous Science: A Two-Year Community Roadmap

arXiv:2607.12113v1 Announce Type: cross Abstract: One year ago, the AISLE roadmap argued that autonomous laboratories operated as isolated islands and proposed a grassroots network organized around fi

safetyarxiv-cs-ai

Towards Self-Evolving Agents: A Human-Inspired Adaptive Exploration-Exploitation Framework for Genetic Network Programming

arXiv:2607.11913v1 Announce Type: cross Abstract: Recent advancements in agentic AI have increasingly moved toward graph-based methods, driven by the demand for explainable, human-centered, and non-li

Towards Vision-Free CIR: Attribute-Augmented Scoring and LLM-Based Reranking for Zero-Shot Composed Image Retrieval

arXiv:2607.12621v1 Announce Type: new Abstract: Recent work has shown that 'Vision-Free'' approaches (representing images as text) can be effective for standard image retrieval tasks. However, it rema

researcharxiv-cs-cv

TRACE: An Operational Reasoning Schema for Auditable Agentic Commitments

arXiv:2607.12480v1 Announce Type: new Abstract: This paper defines TRACE (Typed Reasoning And Commitment Evidence): a typed, versioned schema for recording reasoning traces, a reference procedure for

Traceback Translators Against Forgetting in Continual Fake Speech Detection

arXiv:2607.12569v1 Announce Type: cross Abstract: Fake speech detectors are increasingly challenged by the development of new and more accurate generative models. To cope with this problem, continual

researcharxiv-cs-ai

Applications

TraceSynth: Generating Production-Quality Kernel Traces with Constraint-Guided Diffusion Models

arXiv:2607.12104v1 Announce Type: cross Abstract: Machine learning models for system diagnostics rely on kernel execution traces to capture fine-grained system behavior, but collecting production trac

applicationsarxiv-cs-lg

Tracing Agentic Failure from the Flow of Success

arXiv:2607.12747v1 Announce Type: new Abstract: Failure attribution for LLM-based agentic systems, i.e., identifying which steps in a failure trajectory caused the task to fail, is critical for debugg

agentsarxiv-cs-ai

Track, Rank, Crack: Epistemic Working Memory Scales Multi-Hop Reasoning in Language Agents

arXiv:2607.12267v1 Announce Type: cross Abstract: Language agents that interleave reasoning and tool use degrade sharply as reasoning chains lengthen, even when each individual step is easy. We trace

TRAIL: A Platform for Configurable Human--AI Teaming Experiments

arXiv:2607.12180v1 Announce Type: cross Abstract: An AI teammate's design properties (personality, communication style, when it speaks) can shape a team's trust, coordination, and decisions. Studying

safetyarxiv-cs-ai

Applications

Training against GPT‑Red makes GPT‑5.6 substantially more resilient. To measure this, we replayed some of GPT‑Red’s strongest attacks—none o…

Training against GPT‑Red makes GPT‑5.6 substantially more resilient. To measure this, we replayed some of GPT‑Red’s strongest attacks—none of which our models had seen during training. GPT‑5.6 Sol pro

applicationsopenai--x

Traj-VLN: Learning Pixel-Space Interaction via Autoregressive Trajectory Generation

arXiv:2607.10744v2 Announce Type: replace Abstract: Benefiting from the powerful priors embedded in large-scale pre-training data and the emerging commonsense reasoning ability, large language models

agentsarxiv-cs-cv

Transforming LLMs into Efficient Cross-Encoders via Knowledge Distillation for RAG Reranking

model-releasesarxiv-cs-cl

arXiv:2607.11933v1 Announce Type: new Abstract: Cross-encoders achieve high reranking accuracy in Retrieval-Augmented Generation (RAG) pipelines but impose quadratic inference costs that limit real-ti

Translation as a Computationally Efficient Bridge: Feasibility of English BERT for Low-Resource Languages

arXiv:2607.12612v1 Announce Type: new Abstract: BERT models have revolutionised Natural Language Processing (NLP) through their ability to process unstructured text across diverse domains. However, de

researcharxiv-cs-cl

Trending AND fastest-growing in the same month? Benchmarks change daily, but only @tryramp has the database of real receipts to track this. …

model-releasesfireworks-ai--x

Trending AND fastest-growing in the same month? Benchmarks change daily, but only @tryramp has the database of real receipts to track this. We can confirm: demand for open-weight inference and trainin

TrustVLA: Mechanism-Guided Inference-Time Defense Against Vision-Language-Action Backdoors

arXiv:2607.12571v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models are deployed through pipelines that end users cannot audit, and a poisoned VLA can behave normally on clean observat

safetyarxiv-cs-ro

Local Ai

TSCA-Net: Temporal-Spatial Clique Attention for Interpretable Multimodal Pedestrian Trajectory Prediction

arXiv:2607.11939v1 Announce Type: new Abstract: Accurate pedestrian trajectory prediction in crowded environments remains challenging due to the multimodal uncertainty of human motion and the variable

local-aiarxiv-cs-cv

Tune in today at 10am PST!

Tune in today at 10am PST! Livestream Alert: Run ComfyUI From Claude/Cursor with Comfy MCP Host: @PurzBeats Comfy MCP lets Claude, Cursor, Amp and almost any AI agent you're already using build, run,

model-releasescomfyui--x

Tutorials

UMSS: Towards Unsupervised Multi-modal Semantic Segmentation

arXiv:2607.12372v1 Announce Type: new Abstract: Multimodal semantic segmentation (MSS) is essential for robust perception in complex environments, yet its potential remains largely untapped because of

tutorialsarxiv-cs-cv