All entries

arXiv:2605.12163v1 Announce Type: new Abstract: In language reasoning, longer chains of thought consistently yield better performance, which naturally suggests that visual latent reasoning may likewis

applicationsarxiv-cs-cv

Research

Self-Distilled Trajectory-Aware Boltzmann Modeling: Bridging the Training-Inference Discrepancy in Diffusion Language Models

DGX agent

arXiv:2605.11854v1 Announce Type: new Abstract: Diffusion Language Models (DLMs) have recently emerged as a promising alternative to autoregressive language models, offering stronger global awareness

researcharxiv-cs-cl

13 May 2026

Self-Supervised Laplace Approximation for Bayesian Uncertainty Quantification

model-releasesarxiv-cs-lg

arXiv:2605.12208v1 Announce Type: cross Abstract: Approximate Bayesian inference typically revolves around computing the posterior parameter distribution. In practice, however, the main object of inte

Semi-Supervised Bayesian GANs with Log-Signatures for Uncertainty-Aware Credit Card Fraud Detection

arXiv:2509.00931v3 Announce Type: replace-cross Abstract: We present a novel deep generative semi-supervised framework for credit card fraud detection, formulated as time series classification task. A

SEMIR: Semantic Minor-Induced Representation Learning on Graphs for Visual Segmentation

model-releasesarxiv-cs-cv

arXiv:2605.12389v1 Announce Type: new Abstract: Segmenting small and sparse structures in large-scale images is fundamentally constrained by voxel-level, lattice-bound computation and extreme class im

Local Ai

SenseNova-U1 Technical Report: VAE-free Pixel-level Flow Matching with 32x Compression

local-air-stablediffusion

SenseNova-U1 is a native unified multimodal model built on the NEO-unify architecture that eliminates visual encoders and VAEs, instead using a near-lossless visual interface that preserves semantic s

SenseNova-U1: Unifying Multimodal Understanding and Generation with NEO-unify Architecture

arXiv:2605.12500v1 Announce Type: new Abstract: Recent large vision-language models (VLMs) remain fundamentally constrained by a persistent dichotomy: understanding and generation are treated as disti

agentsarxiv-cs-cv

Sequential Off-Policy Learning with Logarithmic Smoothing

arXiv:2506.10664v2 Announce Type: replace-cross Abstract: Off-policy learning enables training policies from logged interaction data. Most prior work considers the batch setting, where a policy is lea

SEVO: Semantic-Enhanced Virtual Observation for Robust VLA Manipulation via Active Illumination and Data-Centric Collection

arXiv:2605.11114v1 Announce Type: new Abstract: Vision-Language-Action (VLA) and imitation-learning policies trained via community toolchains on low-cost hardware frequently fail when deployed outside

safetyarxiv-cs-ro

ShapeCodeBench: A Renewable Benchmark for Perception-to-Program Reconstruction of Synthetic Shape Scenes

model-releasesarxiv-cs-cv

arXiv:2605.11680v1 Announce Type: new Abstract: We introduce ShapeCodeBench, a synthetic benchmark for perception-to-program reconstruction: given a rendered raster image, a model must emit an executa

Shaping Zero-Shot Coordination via State Blocking

arXiv:2605.11688v1 Announce Type: new Abstract: Zero-shot coordination (ZSC) aims to enable agents to cooperate with independently trained partners without prior interaction, a key requirement for rea

agentsarxiv-cs-lg

Shapley Value Approximation Based on k-Additive Games

arXiv:2502.04763v2 Announce Type: replace-cross Abstract: The Shapley value is the prevalent solution for fair division problems in which a payout is to be divided among multiple agents. By adopting a

ShardTensor: Domain Parallelism for Scientific Machine Learning

arXiv:2605.11111v1 Announce Type: cross Abstract: Scientific Machine Learning (SciML) faces unique challenges for extreme-resolution data, with mitigations that often fail to scale or degrade the accu

Tutorials

Sharpen Your Flow: Sharpness-Aware Sampling for Flow Matching

arXiv:2605.11547v1 Announce Type: new Abstract: Flow matching models generate samples by numerically integrating a learned velocity field, with each integration step requiring a neural network evaluat

tutorialsarxiv-cs-lg

SI-Diff: A Framework for Learning Search and High-Precision Insertion with a Force-Domain Diffusion Policy

arXiv:2605.12247v1 Announce Type: new Abstract: Contact-rich assembly is fundamental in robotics but poses significant challenges due to uncertainties in relative poses, such as misalignments and smal

safetyarxiv-cs-ro

Applications

Sign Language Recognition and Translation for Low-Resource Languages: Challenges and Pathways Forward

arXiv:2605.12096v1 Announce Type: new Abstract: Sign languages are natural, visual-gestural languages used by Deaf communities worldwide. Over 300 distinct sign languages remain severely low-resource

applicationsarxiv-cs-cl

Sign up below for a Hermes Agent meetup in Montreal!

Sign up below for a Hermes Agent meetup in Montreal! Save the date for May 29 at 6:30pm. Be here in Montreal or be square. You are cordially invited to a cozy apartment MEETUP hosted by myself and @vn

agentsnous-research--x

Simpson's Paradox in Behavioral Curves: How Aggregation Distorts Parametric Models of User Dynamics

arXiv:2605.11017v1 Announce Type: new Abstract: Behavioral curve modeling -- fitting parametric functions to engagement-versus-exposure data -- is standard practice in recommendation, advertising, and

Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptation

arXiv:2603.15759v2 Announce Type: replace-cross Abstract: Robot learning requires adaptation methods that improve reliably from limited, mixed-quality interaction data. This is especially challenging

Simulation-Ready Cluttered Scene Estimation via Physics-aware Joint Shape and Pose Optimization

arXiv:2602.20150v2 Announce Type: replace-cross Abstract: Estimating simulation-ready scenes from real-world observations is crucial for downstream planning and policy learning tasks. Regretfully, exi

safetyarxiv-cs-cv

Single-Shot HDR Recovery via a Video Diffusion Prior

arXiv:2605.11628v1 Announce Type: new Abstract: Recent generative methods for single-shot high dynamic range (HDR) image reconstruction show promising results, but often struggle with preserving fidel

researcharxiv-cs-cv

SkillGen: Verified Inference-Time Agent Skill Synthesis

arXiv:2605.10999v1 Announce Type: new Abstract: Skills are a promising way to improve LLM agent capabilities without retraining, while keeping the added procedure reusable and controllable. However, h

agentsarxiv-cs-lg

SkillGraph: Skill-Augmented Reinforcement Learning for Agents via Evolving Skill Graphs

arXiv:2605.12039v1 Announce Type: new Abstract: Skill libraries enable large language model agents to reuse experience from past interactions, but most existing libraries store skills as isolated entr

safetyarxiv-cs-cl

SkillSafetyBench: Evaluating Agent Safety under Skill-Facing Attack Surfaces

model-releasesarxiv-cs-cl

arXiv:2605.12015v1 Announce Type: cross Abstract: Reusable skills are becoming a common interface for extending large language model agents, packaging procedural guidance with access to files, tools,

Slicing and Dicing: Configuring Optimal Mixtures of Experts

model-releasesarxiv-cs-cl

arXiv:2605.11689v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) architectures have become standard in large language models, yet many of their core design choices - expert count, granularit

Smart moves: Building resilient transportation systems with Google AI

What does transportation mean to you? For some, it’s making sure the train is on schedule so they can get to work on time. Maybe it’s making sure you have time connecting between flights. Maybe it’s a

safetygoogle-cloud-ai

Smooth-Rigid-Body Contact as a ReLCP: A Recursively Generated Linear Complementarity Problem

arXiv:2506.14097v2 Announce Type: replace Abstract: This paper reformulates complementarity-based time-stepping for frictionless nonsmooth contact between smooth rigid bodies as a recursively generate

researcharxiv-cs-ro

Tutorials

Smoothed Analysis of Learning from Positive Samples

arXiv:2504.10428v2 Announce Type: replace-cross Abstract: Binary classification from positive-only samples is a variant of PAC learning where the learner receives i.i.d. positive samples and aims to l

tutorialsarxiv-cs-lg

Tutorials

Smoothness Errors in Dynamics Models and How to Avoid Them

arXiv:2602.05352v2 Announce Type: replace Abstract: Modern neural networks have shown promise for solving partial differential equations over surfaces, often by discretizing the surface as a mesh and

tutorialsarxiv-cs-lg

so cool to see another blazing fast database built on vortex!

so cool to see another blazing fast database built on vortex! Just announced at Interrupt! SmithDB. Agent traces have outgrown the databases built to hold them. That’s why we built SmithDB, a purpose-

agentsharrison-chase--x

So many agent builders at Langchain Interrupt 2026 😍😍😍

At LangChain Interrupt 2026, multiple agent builder tools and frameworks were showcased, reflecting the growing ecosystem of agentic AI development. This post highlights the proliferation of agent-bui

agentsharrison-chase--x

so we built psql_bm25s. exact BM25 retrieval. native Postgres access method. ~23x faster than pg_search on the standard benchmark. retrieval…

model-releasesemad-mostaque--x

so we built psql_bm25s. exact BM25 retrieval. native Postgres access method. ~23x faster than pg_search on the standard benchmark. retrieval stops being a budget item. the harness stops rationing. the

SOAR: Regression-based LiDAR Relocalization for UAVs

model-releasesarxiv-cs-cv

arXiv:2602.13267v3 Announce Type: replace Abstract: Regression-based LiDAR relocalization has recently emerged as a promising solution for high-precision positioning in GNSS-denied environments. Howev

SOAR: Scale Optimization for Accurate Reconstruction in NVFP4 Quantization

arXiv:2605.12245v1 Announce Type: new Abstract: NVFP4 has recently emerged as an efficient 4-bit microscaling format for large language models (LLMs), offering superior numerical fidelity with native

Sobolev Regularized MMD Gradient Flow

arXiv:2605.11884v1 Announce Type: new Abstract: We propose Sobolev-regularized Maximum Mean Discrepancy (SrMMD) gradient flow, a regularized variant of maximum mean discrepancy (MMD) gradient flow bas

SoftBank reports Q4 net income up 3x to ~11.6B, above ~1.5B est., driven by a 25B gain in its OpenAI stake; by October, its OpenAI investment will hit 64.6B (David Keohane/Financial Times)

David Keohane / Financial Times: SoftBank reports Q4 net income up 3x to ~11.6B, above ~1.5B est., driven by a 25B gain in its OpenAI stake; by October, its OpenAI investment will hit 64.6B — Japanese

SoK: Unlearnability and Unlearning for Model Dememorization

arXiv:2605.11592v1 Announce Type: new Abstract: Advanced model dememorization methods, including availability poisoning (unlearnability) and machine unlearning, are emerging as key safeguards against

Solve the Loop: Attractor Models for Language and Reasoning

model-releasesarxiv-cs-cl

arXiv:2605.12466v1 Announce Type: cross Abstract: Looped Transformers offer a promising alternative to purely feed-forward computation by iteratively refining latent representations, improving languag

Local Ai

SOMA: Efficient Multi-turn LLM Serving via Small Language Model

arXiv:2605.11317v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in multi-turn dialogue settings where preserving conversational context across turns is essential

local-aiarxiv-cs-cl

Source: Cerebras priced its IPO at 185 per share, above the expected range of 150 to 160, raising at least 5.55B and valuing it at $56.4B fully diluted (Jordan Novet/CNBC)

Jordan Novet / CNBC: Source: Cerebras priced its IPO at 185 per share, above the expected range of 150 to 160, raising at least 5.55B and valuing it at 56.4B fully diluted — Cerebras Systems, a maker

Applications

Sources: Apollo Global, Morgan Stanley, and others are testing Grok internally as part of xAI's finance sector push, but financiers are rarely using it for work (Carmen Arroyo/Bloomberg)

Carmen Arroyo / Bloomberg: Sources: Apollo Global, Morgan Stanley, and others are testing Grok internally as part of xAI's finance sector push, but financiers are rarely using it for work — Elon Musk'

applicationstechmeme

Sources: Arm and SoftBank expressed preliminary interest in acquiring Cerebras weeks before its expected IPO; Cerebras rebuffed them (Bloomberg)

Bloomberg: Sources: Arm and SoftBank expressed preliminary interest in acquiring Cerebras weeks before its expected IPO; Cerebras rebuffed them — Arm Holdings Plc and its majority owner SoftBank Group

Sources: Google plans to announce a new Gemini model at its I/O conference next week; the model will land roughly in the class of GPT-5.5, but short of Mythos (Alex Heath/Sources)

Alex Heath / Sources: Sources: Google plans to announce a new Gemini model at its I/O conference next week; the model will land roughly in the class of GPT-5.5, but short of Mythos — Sources say that

model-releasestechmeme

Sources: Microsoft is in discussions to acquire LLM developer Inception; SpaceX also courted Inception, which is looking for a price of over $1B (Reuters)

Reuters: Sources: Microsoft is in discussions to acquire LLM developer Inception; SpaceX also courted Inception, which is looking for a price of over $1B — Microsoft (MSFT.O) is shopping for artificia

Sources: Mistral has been developing a cybersecurity-focused AI model and held discussions about it with European banks, which don't have access to Mythos (Bloomberg)

Bloomberg: Sources: Mistral has been developing a cybersecurity-focused AI model and held discussions about it with European banks, which don't have access to Mythos — French artificial intelligence s

model-releasestechmeme

Space Syntax-guided Post-training for Residential Floor Plan Generation

arXiv:2602.22507v2 Announce Type: replace-cross Abstract: Residential floor plan generation requires not only geometric fidelity but also spatial configurational logic: shared living spaces should be

safetyarxiv-cs-cv

Sparse Attention Remapping with Clustering for Efficient LLM Decoding on PIM

arXiv:2505.05772v2 Announce Type: replace Abstract: Transformer-based models are the foundation of modern machine learning, but their execution, particularly during autoregressive decoding in large la

researcharxiv-cs-cl

Sparse Offline Reinforcement Learning with Corruption Robustness

arXiv:2512.24768v3 Announce Type: replace-cross Abstract: We investigate robustness to strong data corruption in offline sparse reinforcement learning (RL). In our setting, an adversary may arbitraril