Search: “inference”

From Diagnosis to Correction: Benchmarking and Improving Real-World Table Parsing

arXiv:2608.09842v1 Announce Type: new Abstract: Recent document parsers achieve table TEDS scores above 93 on OmniDocBench v1.6, yet community feedback and our audit reveal persistent failures on comp

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

From Rebound to Remedy: Understanding and Mitigating Reward Hacking via Representation Engineering

DGX agent

arXiv:2604.01476v2 Announce Type: replace-cross Abstract: Reinforcement learning for LLMs is vulnerable to reward hacking, where models exploit shortcuts to maximize reward without solving the intende

safetyarxiv-cs-cl

11 Aug 2026

From Single Chatbots to Governed Agent Ecosystems: An Agentic AI Pattern Catalogue and Orchestration Framework for Mission-Critical Hospital Information Management Systems

arXiv:2608.07627v1 Announce Type: new Abstract: Hospitals are racing to embed AI, while coping with the surge in adaptation of the technology in other industries, into the triage management, documenta

safetyarxiv-cs-ai

GenTrack3: Hybrid Stochastic-Deterministic Online Multi-Object Tracking with Cluster-Aware Association

arXiv:2608.09581v1 Announce Type: new Abstract: Multi-object tracking (MOT) involves maintaining consistent target identities as objects dynamically enter and leave a scene. Deterministic approaches,

researcharxiv-cs-cv

Geometry Beats Estimated Depth: RGB-Only Multi-Camera 3D Tracking under Sim2Real

arXiv:2608.07579v1 Announce Type: cross Abstract: The AI City Challenge 2026 Track 1 evaluates multi-camera 3D perception in large indoor warehouses under a synthetic-to-real (Sim2Real) setting; depth

Had so much fun giving this talk at @sequoia about @harvey’s moneyball approach to building a research lab. The biggest mistake I made in th…

Had so much fun giving this talk at @sequoia about @harvey’s moneyball approach to building a research lab. The biggest mistake I made in the early days of Harvey was trying to play the Yankees baseba

agentssonya-huang--x

How does OpenAI move so quickly? I'm the first to report on friction@openai.com—the AI age's version of a Jeff Bezos question mark email. He…

How does OpenAI move so quickly? I'm the first to report on friction@openai.com—the AI age's version of a Jeff Bezos question mark email. Here's how it works. https://fortune.com/2026/08/11/openai-emp

agentsyohei-nakajima--x

How Much Does It Cost to Answer My Question? Benchmarking Cloud VLM-based VQA Systems

arXiv:2608.07861v1 Announce Type: new Abstract: Vision-language models (VLMs) are becoming a practical backend for mobile visual question answering (VQA) systems, enabling smartphones and smart glasse

HSMLA: Hierarchical Softmax Multi-scale Linear Attention for Efficient Vision Transformers

arXiv:2608.07616v1 Announce Type: new Abstract: Vision transformers face significant computational overheads in high-resolution dense prediction due to the quadratic complexity of self-attention. Line

local-aiarxiv-cs-cv

Imaginative Generative AI: Crossing the Entropy Wall into Worlds Beyond Imitation

arXiv:2608.09385v1 Announce Type: cross Abstract: Generative AI models are primarily designed to imitate the data distribution, an objective that neither corrects diversity lost by a learned generator

In-Context Density Estimation for Tabular Data

arXiv:2608.09348v1 Announce Type: new Abstract: Density estimation underlies many unsupervised tasks on tabular data such as anomaly detection, out-of-distribution detection, and data augmentation. Al

researcharxiv-cs-lg

InfMem: Learning System-2 Memory Control for Long-Context Agent

arXiv:2602.02704v2 Announce Type: replace Abstract: Reasoning over ultra-long documents requires synthesizing sparse evidence scattered across distant segments under strict memory constraints. While s

agentsarxiv-cs-cl

Integrated Multimodal AI System for Retrieval-Augmented Reasoning, Object Sensing, and Damage Analysis

arXiv:2608.08935v1 Announce Type: new Abstract: This work presents a unified multimodal AI system for damage assessment that integrates retrieval-augmented generation (RAG) models, thermal spectrum pe

local-aiarxiv-cs-ai

Knowing You Is Everything: LLM Agents Achieve Near-Perfect Profile-Consistent Reaction Prediction in Social Media Simulation

arXiv:2608.07498v1 Announce Type: cross Abstract: Autonomous AI agents in social media present concrete risks to democratic discourse and platform governance, while also offering tools for pre-deploym

Learning Structural Illumination for Unsupervised Low-light Enhancement

arXiv:2608.08153v1 Announce Type: new Abstract: Existing unsupervised low-light image enhancement (LLIE) methods often estimate illumination directly from the entire low-light input, without separatin

LexKairos: Benchmarking Legal Temporal Capabilities in LLMs

model-releasesarxiv-cs-cl

arXiv:2608.09106v1 Announce Type: new Abstract: Large language models (LLMs) have demonstrated strong performance across a wide range of legal tasks. In legal practice, time is a critical concept that

LIBAD: A Multimodal Anomaly Detection Benchmark for Li-Ion Battery Electrode Manufacturing

arXiv:2608.07958v1 Announce Type: new Abstract: Multimodal industrial anomaly detection has largely focused on discrete products using strongly correlated RGB and 3D observations, leaving continuous p

LITEWAY: LIghtweight HAR via Temporal Efficient highWAY

arXiv:2608.09421v1 Announce Type: cross Abstract: Wearable human activity recognition (HAR) remains challenging due to the computational and energy constraints of deep learning models on resource-limi

[llama.cpp PR #26608] Ling-3.0 support (unmerged)

model-releasesr-localllama

aetherbird has done some great work getting Ling-3.0 to work in llama.cpp. The architecture is generally identical to deepseekv2. I recently added a microscopic 40 line PR to his that adds support for

LLM-Driven AutoML for Cross-Lingual Handwritten OCR: Closed-Loop Neural Architecture Search with GPT-5, GPT-4o, and Claude Sonnet 4

arXiv:2607.15509v2 Announce Type: replace-cross Abstract: We present a fully automated closed-loop AutoML framework that uses GPT-5, GPT-4o, and Claude Sonnet 4 as autonomous neural architecture desig

Matryoshka Language Model Suites

arXiv:2608.09703v1 Announce Type: new Abstract: Training a language model suite classically requires training each model separately and serving them independently. We improve both training and inferen

Measuring the Wrong Thing: Internal Harmfulness Scores Anti-Rank Successful Jailbreaks

arXiv:2608.09624v1 Announce Type: cross Abstract: Internal safety scores judge a prompt before any text is generated, and they are validated by how well they separate harmful prompts from benign ones.

Mind the Hook: Source-Level Auditing of Privacy Defenses in Retrieval-Augmented Generation

model-releasesarxiv-cs-lg

arXiv:2608.09001v1 Announce Type: cross Abstract: Black-box privacy scores for retrieval-augmented generation (RAG) are difficult to interpret unless the audited defense's active pipeline hook is know

MiniMax-H3: ~38 GB less VRAM with Runtime LoRA Bypass — DoRA Dynamic LoRA Loader v1.0.39

local-air-stablediffusion

GitHub: https://github.com/xmarre/ComfyUI-DoRA-Dynamic-LoRA-Loader Release v1.0.39: https://github.com/xmarre/ComfyUI-DoRA-Dynamic-LoRA-Loader/releases/tag/v1.0.39 Also available through ComfyUI Manag

Mitigating Gender Bias in English to Romanian Machine Translation

arXiv:2608.08606v1 Announce Type: cross Abstract: Machine translation (MT) systems often fail to correctly translate gender, especially when converting from a gender-neutral language like English to a

safetyarxiv-cs-ai

Mitigating Over-Personalization in LLMs via Structured Memory

arXiv:2608.08300v1 Announce Type: new Abstract: Conversational assistants increasingly rely on persistent long-term memory to personalize responses across sessions. However, when stored user informati

Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models

arXiv:2608.09696v1 Announce Type: new Abstract: Predicting the answer to interventional ``what if'' questions --- the outcome of an action never taken --- requires a mechanistic, causal model, not a c

MoE-Prism: Disentangling Monolithic Experts for Elastic MoE Services via Model-System Co-Designs

arXiv:2510.19366v2 Announce Type: replace Abstract: Mixture-of-Experts (MoE) scales model capacity through sparse activation, and is becoming an important architecture for large language models (LLMs)

researcharxiv-cs-cl

Motif 3: Technical Report

arXiv:2608.09119v1 Announce Type: new Abstract: We introduce Motif 3, a decoder-only Mixture-of-Experts language model with 314 billion total parameters and 13.2 billion activated per token. Each spar

safetyarxiv-cs-ai

MRBench: A Comprehensive Benchmark for Human Motion-Text Retrieval

arXiv:2608.07993v1 Announce Type: new Abstract: Human motion-text retrieval provides a rigorous means of assessing cross-modal alignment. Prevailing benchmarks are dominated by homogeneous indoor moti

Multi-Task Consistency-based Detection of Adversarial Attacks

arXiv:2608.07750v1 Announce Type: cross Abstract: Deep Neural Networks (DNNs) have found successful deployment in numerous vision perception systems. However, their susceptibility to adversarial attac

agentsarxiv-cs-ai

Multimodal Skin Lesion Classification with Swin Transformer and Clinical Metadata Fusion

arXiv:2608.07574v1 Announce Type: new Abstract: Skin lesion classification plays an important role in supporting the early diagnosis of skin cancer. However, automated analysis remains challenging due

researcharxiv-cs-cv

Nemotron 3.5 Lightning is available in LM Studio! The model is 30B MoE (3B active), can run very fast, and is trained for high volume agenti…

model-releaseslm-studio--x

Nemotron 3.5 Lightning is available in LM Studio! The model is 30B MoE (3B active), can run very fast, and is trained for high volume agentic use cases. Model page: https://lmstudio.ai/models/nvidia/n

Not All Visual Tokens Are Equally Safe to Remove:Consequence-Sensitive Visual Token Compression

arXiv:2608.09176v1 Announce Type: cross Abstract: Visual token compression for vision--language models (VLMs) has largely relied on criteria such as attention, redundancy, and uncertainty to maximize

Hardware

NVIDIA JetPack 7.2.1 Adds Agentic Video Skills and T3000 Emulation

NVIDIA JetPack 7.2.1 adds agentic video skills with the unified jetson‑videosdk, allowing programmable, device-aware video workflows that link developer intent to live device discovery and performance

hardwarenvidia-developer

NVIDIA Nemotron 3.5 Lightning is now live on Together AI. The fastest open model in its class is built for always-on agents that need to com…

model-releasestogether-ai--x

NVIDIA’s Nemotron 3.5 Lightning—a fast open AI model designed for always‑on agents that perform high‑volume, specialized work—has been launched on the Together AI platform as of 11 August 2026. NVIDIA

OD-Gear: Online Decomposition and Group Sampling for Expert-Guided Adversarial Routing in Scalable Capacitated Vehicle Routing

arXiv:2602.00488v3 Announce Type: replace Abstract: Solving large-scale capacitated vehicle routing problems (CVRP) is hindered by the high complexity of classical heuristics and the limited generaliz

safetyarxiv-cs-lg

On-Device Multi-Species Malaria Detection with Uncertainty-Calibrated Slide-Level Aggregation

arXiv:2608.08566v1 Announce Type: cross Abstract: Malaria remains a leading cause of mortality in resource-limited settings, where expert microscopists are scarce. Automated diagnosis based on microsc

local-aiarxiv-cs-ai

Opportunity Is Not Realizability: Selection-Valid Diagnostics for Multi-LLM Routing

arXiv:2608.08265v1 Announce Type: new Abstract: Oracle routing measures how much a pool of language models could gain from per-query selection, but the diagnostic has two flaws: testing against a best

researcharxiv-cs-lg

Overcoming Data Scarcity and Confidentiality in Hardware Assurance via Synthetic Generation

arXiv:2608.09914v1 Announce Type: cross Abstract: Hardware assurance relies on scanning electron microscopy (SEM) to verify nanoscale structures, but assembling the large, high-quality datasets requir

researcharxiv-cs-cv

PAM: Training Policy-Aligned Moderation Filters at Scale

arXiv:2505.19766v4 Announce Type: replace Abstract: Large language models (LLMs) remain vulnerable to misalignment and jailbreaks, making external safeguards like moderation filters essential, yet exi

safetyarxiv-cs-cl

PolypSteer: Counterfactual Endoscopic Synthesis via Training-Free Activation Steering

arXiv:2603.07066v2 Announce Type: replace-cross Abstract: Generative diffusion models are increasingly used for medical imaging data augmentation, but text prompting cannot produce causal training dat

Population-Level Generative Modeling for Ranking Data

model-releasesarxiv-cs-lg

arXiv:2608.08422v1 Announce Type: cross Abstract: Ranking data arise in scientific and machine learning applications, including recommendation systems, information retrieval, voting, marketing, and AI

Position Bias in Ordinal Classification: A Systematic Evaluation

arXiv:2608.08869v1 Announce Type: new Abstract: Large language models are increasingly used for ordinal classification, yet semantically equivalent changes to prompt organization can alter their predi

safetyarxiv-cs-cl

Preserving Item Semantics for Free: Rethinking Token Initialization in LLM-Based Generative Recommendation

model-releasesarxiv-cs-lg

arXiv:2608.07816v1 Announce Type: cross Abstract: Recent advances in generative recommendation (GR) leverage large language models (LLMs) as recommender backbones, enabling LLMs to directly generate r

Probabilistic Circuits for Knowledge Graph Completion with Reduced Rule Sets

arXiv:2508.06706v2 Announce Type: replace Abstract: Rule-based methods for knowledge graph completion provide explainable results, but often require tens of thousands of rules to achieve competitive p

Applications

Progressive Learned Image Compression for Machine Perception

arXiv:2512.20070v2 Announce Type: replace Abstract: Recent advances in learned image codecs have extended from human perception toward machine perception However, progressive image compression with fi

applicationsarxiv-cs-cv

RAGMesh with FaME-G2E: Long-Form Text-Driven 3D Face Generation and Editing

arXiv:2608.09186v1 Announce Type: new Abstract: Text-driven 3D face generation and editing remains challenging due to the difficulty of translating long-form descriptions into fine-grained facial geom