All entries

ASTRA: ADMM-Accelerated Topology Reconfiguration for Dynamic Satellite Constellations

arXiv:2609.33993v1 Announce Type: new Abstract: Dynamic topology reconfiguration is central to the reliability and efficiency of large satellite constellations, yet many existing approaches rely on id

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Agents

Attention-based Hierarchical Variational Information Bottleneck for Robust Multi-Agent Communication under Variable Bandwidth

DGX agent

arXiv:2609.34860v1 Announce Type: new Abstract: Learning-based multi-agent communication under limited bandwidth does not only require deciding what to communicate, but also structuring messages so th

agentsarxiv-cs-lg

29 Sept 2026

Attention Graphons: A Graph Limit Perspective on Graph Transformers

arXiv:2609.35620v1 Announce Type: new Abstract: Graph Transformers produce, for each attention head, a dense nimes n matrix of learned pairwise interactions. We ask a fundamental question: do these at

AttentionViG: Cross-Attention-Based Dynamic Neighbor Aggregation in Vision GNNs

arXiv:2509.25570v2 Announce Type: replace Abstract: Vision Graph Neural Networks (ViGs) have demonstrated promising performance in image recognition tasks against Convolutional Neural Networks (CNNs)

Attribution Gaps in Zero-Training LLM+OVOD Pipelines: A Fine-Grained Analysis of the CAAP--SNAP Discrepancy

arXiv:2609.32567v1 Announce Type: new Abstract: LAOD and similar zero-training LLM+open-vocabulary-detector (OVOD) pipelines score two things separately: class-agnostic localization accuracy (CAAP) an

Safety

Attribution Without a Second Pass: Inline Per-Sample Gradient Provenance at ~1% Overhead

arXiv:2609.32380v1 Announce Type: new Abstract: Data attribution methods used in practice (TRAK, LoGRA, EK-FAC) are post-hoc: after training they make a second pass over the training set to recompute

safetyarxiv-cs-lg

Audio-JEPA: Joint-Embedding Predictive Architecture for Audio Representation Learning

arXiv:2507.02915v2 Announce Type: replace-cross Abstract: Building on the Joint-Embedding Predictive Architecture (JEPA) paradigm, a recent self-supervised learning framework that predicts latent repr

Safety

Audit-First VAPO: Risk-Certified Selective Updates under Imperfect Verification

arXiv:2609.33662v1 Announce Type: cross Abstract: Imperfect verifiers can assign a harmful update direction even when clipping and regularization bound its magnitude. We introduce Audit-First VAPO, wh

safetyarxiv-cs-lg

Agents

Audit the Scaffold, Not the Checkpoint: A Stationarity Dichotomy for Recursive Self-Improvement in Agentic Coding

arXiv:2609.34924v1 Announce Type: new Abstract: An auditor who checks whether a system's weights are frozen is checking the wrong thing. Our stationarity dichotomy says that iterative self-modificatio

agentsarxiv-cs-lg

Auditing Information Disclosure During Large-Scale Gradient-Based Training via Gradient Uniqueness

arXiv:2510.10902v3 Announce Type: replace Abstract: Auditing information disclosure across every datapoint during the training of LLMs is challenging. We propose a principled, attack-agnostic approach

Auditing Quality Filters for Long-Tail Human Data Curation

arXiv:2609.31896v1 Announce Type: new Abstract: Robots on construction sites must detect workers who are kneeling or bending, which we call low poses. These workers can be lost from training datasets

Augmented Feature Boosting for Multicalibration

arXiv:2609.33891v1 Announce Type: new Abstract: Multicalibration requires a predictor's residuals to be unbiased not only globally, but also after conditioning on the predictor's own level sets and re

Tutorials

Augmenting Visual Anomaly Detection with Automated Interpretability

arXiv:2609.33818v1 Announce Type: new Abstract: Visual anomaly detectors identify deviations from known-normal data, but their anomaly signals may mix evidence of actual anomalies with benign visual v

tutorialsarxiv-cs-cv

AutoExpert: Automating 3D LiDAR Annotation from Expert-Crafted Guidelines

arXiv:2506.02914v3 Announce Type: replace Abstract: The contemporary paradigm of scaling data annotation, crucial for developing machine learning solutions, is to hire ordinary human annotators and in

AutoHGNN: Robust and Efficient Neural Architecture Search for Hypergraph Neural Networks

arXiv:2609.33392v1 Announce Type: new Abstract: Hypergraph neural networks have achieved significant success in recent years. However, manual architecture crafting is labor-intensive and often fails t

Automated Lesion Segmentation of Stroke MRI Using nnU-Net: A Comprehensive External Validation Across Acute and Chronic Lesions

arXiv:2601.08701v2 Announce Type: replace-cross Abstract: Accurate and generalisable segmentation of stroke lesions from magnetic resonance imaging (MRI) is essential for advancing clinical research,

Automated Species Identification in Camera Trap Images for Wildlife Conservation

arXiv:2609.35420v1 Announce Type: new Abstract: Wildlife conservation involves protecting, preserving, and managing wildlife species and their habitats. With today's rapid pace of human development, c

Autonomous phase discovery

arXiv:2609.33802v1 Announce Type: cross Abstract: Understanding quantum phases of matter has long relied on physicists' intuition and mathematical tools such as symmetry and topology. Remarkably succe

Autonomous Research Project Management as an Agent Skill: A Case Study in Exact Spectral Spatial Regression

arXiv:2609.31683v1 Announce Type: new Abstract: This work presents an end-to-end demonstration of autonomous machine learning research conducted by an agent skill on consumer hardware. The demonstrati

AutoRef: Harness Optimization for Agentic Multi-Reference Image Generation

arXiv:2609.35530v1 Announce Type: new Abstract: Recent image generation models can take multiple reference images as input and combine them into a new image. However, multi-reference image generation

Applications

AutoSDT: Scaling Data-Driven Discovery Tasks Toward Open Co-Scientists

arXiv:2506.08140v2 Announce Type: replace Abstract: Despite long-standing efforts in accelerating scientific discovery with AI, building AI co-scientists remains challenging due to limited high-qualit

applicationsarxiv-cs-lg

Averaged Mirror Descent and Dual Gradient Methods: Convergent Algorithms for Entropic Gromov-Wasserstein Problem

arXiv:2609.31848v1 Announce Type: new Abstract: The Gromov-Wasserstein (GW) distance measures the discrepancy between metric measure (mm) spaces and identifies optimal alignments between them based so

Awaken, Then Scale: Tiny Adaptation for MRI Reconstruction

arXiv:2609.31813v1 Announce Type: cross Abstract: Dormant Awakening (DA) specializes a magnetic resonance imaging (MRI) reconstruction model by fitting a small set of zero weights to one labeled slice

b11243

ci : update the oneAPI toolkit to 2026.1 (#29273) ci : update the oneAPI toolkit to 2026.1 oneDNN is removed from Intel Deep Learning Essentials in 2026.0, so staying on the deep-learning-essentials p

b11245

common : use fs::path for cache dirs (#29595) Avoid useless string conversions on Windows. No need for BSD or emscripten special cases. Signed-off-by: Adrien Gallouët angt@huggingface.co Website: http

b11246

chat : fix Muse Glimmer ignoring response_format json_schema with --jinja (#29615) chat : fix Muse Glimmer ignoring response_format json_schema with --jinja Fixes #29613 chat : accept json fences and

b11247

vulkan : reuse descriptor sets when bindings are constant (#29280) vulkan : reuse descriptor sets when bindings are constant vulkan : bump buffer_destroy_count before destroying the buffer Website: ht

b11249

llama : fix init in several tools/examples (#29632) Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/51066139 macOS/iOS: macOS Apple Silicon (arm64) macOS Ap

b11254

ggml : collect all input tensors into graph_inputs (#29634) graph_inputs was populated while splitting the graph, so it only contained the inputs that are used as srcs of some node. With pipeline para

b11255

common : add fs_write_atomic() (#29642) Check for buffered write errors when closing downloaded files. Use UTF-8 paths when writing ETag files on Windows. Write in binary mode on Windows. Signed-off-b

b11256

tests : adjust server string regex to also match m2 utlra results (#29648) Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/51122498 macOS/iOS: macOS Apple S

b11257

common : use fs::path for config dir (#29649) Signed-off-by: Adrien Gallouët angt@huggingface.co Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/51176378 ma

b11258

server : remove the built-in UI's service worker when the UI is not served (#29565) With --path or --no-ui, /sw.js returned 404, and a 404 does not remove a service worker, so browsers kept showing th

b11259

common : stop accepting draft tokens at EOG (#29638) common : stop accepting draft tokens at EOG cont : remove the test Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/a

b11260

hexagon: add FP32 GELU_ERF and GEGLU_ERF support (#29631) hexagon: add FP32 GELU_ERF and GEGLU_ERF support hex-erf: reduce register pressure in kernels Website: https://llama.app Attestations: https:/

b11261

ggml : require input tensors to be GGML_OP_NONE (#29647) Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/51266318 macOS/iOS: macOS Apple Silicon (arm64) mac

b11262

ggml : accumulate f16 dot products in f32 on AVX512-FP16 (#29545) Supersedes #29530 Signed-off-by: Adrien Gallouët angt@huggingface.co Website: https://llama.app Attestations: https://github.com/ggml-

b11263

vocab : keep NORMAL in PLaMo-2 and PLaMo-3 (#29580) vocab : keep NORMAL in PLaMo-2 and PLaMo-3 The PLaMo-2 and PLaMo-3 vocabularies mark as NORMAL. Current EOG token heuristic matched it by text and a

b11264

ggml : fix c++ odr by properly using GGML_COMMON_DECL_CPP (#29504) fix c++ odr by properly using GGML_COMMON_DECL_CPP using actual field rather than macro Co-authored-by: Georgi Gerganov ggerganov@gma

b11265

vulkan: MOE aware mat_mul_id tile selection (#29182) mut_mul_id selected its matmul tile with total token count. For MoE dispatch grid the true N per workgroup is per-expert rows. At pp128 on Sarvam 3

b11266

vulkan : Load F32 A matrix 2 at a time when its 2-aligned (#29254) It turns out Intel doesn't particularly like loading F32s one at a time and we already have the _2aliagned load logic in mul_mat_vec,

BA-DPO: Bias-Adjusted Direct Preference Optimization for Language Model Alignment

arXiv:2609.35044v1 Announce Type: new Abstract: Preference-based alignment methods such as Direct Preference Optimization (DPO) use pairwise preferences labeled by human annotators to fine-tune langua

Back-Tracking from Clarity: Self-Learning to See Text from Afar

arXiv:2609.32477v1 Announce Type: new Abstract: We propose a self-supervised framework designed to enhance the capability of scene text detectors in identifying and recognizing text in scenarios where

Backdoor as Probe: Test-Time Adversarial Defense for CLIP

arXiv:2609.34641v1 Announce Type: new Abstract: Test-time adversarial defense improves the robustness of vision-language foundation models such as CLIP without retraining. However, adversarial activat

Background Gradients Shape Memorization in Flow Matching

arXiv:2609.33210v1 Announce Type: new Abstract: Repetition is closely associated with memorization in generative models, but how other training images affect the retention and copying of targets remai

Balancing Early Performance Sacrifices with Long-Term Gains: Scaling Learning-Rate Warmup Duration Across Training Horizons

arXiv:2609.33041v1 Announce Type: new Abstract: Learning-rate warmup is a standard technique in language-model training, yet its duration remains largely heuristic. Common approaches use either a fixe

BALAR : A Bayesian Agentic Loop for Active Reasoning

arXiv:2605.05386v2 Announce Type: replace-cross Abstract: Large language models increasingly operate in interactive settings where solving a task requires multiple rounds of information exchange with

Applications

Bayesian Active Learning for Intent Disambiguation in Interactive Robot Planning

arXiv:2609.34270v1 Announce Type: new Abstract: Interactive robot planning requires robots to infer and execute human intentions from natural language instructions that are often ambiguous, incomplete

applicationsarxiv-cs-ro