Search: “inference”

Beyond Token Positions: Safety Alignment Across Denoising Steps in Diffusion Language Models

arXiv:2609.00495v1 Announce Type: cross Abstract: Diffusion large language models (dLLMs) generate text through iterative denoising rather than left-to-right decoding. This generation paradigm introdu

Hardware

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

CacheBridge: Efficient Cross-Model KV Cache Transfer

DGX agent

arXiv:2609.00891v1 Announce Type: new Abstract: Sharing context between LLMs in a multi-model system requires the receiving model to prefill the shared prefix because KV caches are model-specific. Rec

hardwarearxiv-cs-ai

2 Sept 2026

Commit-first LLM judging inherits the judge's own errors

arXiv:2609.00088v1 Announce Type: cross Abstract: LLM judges, models that score another system's output, can be gamed by the systems they score. Recent work identifies one defence that works: the judg

Conditional Flow Matching for Cross-Field MRI Harmonisation

model-releasesarxiv-cs-cv

arXiv:2609.00960v1 Announce Type: new Abstract: Magnetic resonance images of the same subject look markedly different across field strengths, which complicates the comparison and pooling of data acros

Connectivity-Aware Graph Extension for Decentralized Multi-Robot Exploration

arXiv:2609.00804v1 Announce Type: new Abstract: Exploring unknown environments with multiple UAVs requires coordination under intermittent communication, making decentralized operation a baseline assu

researcharxiv-cs-ro

Consistency Without Alignment: Item-Sensitive Language Models Indistinguishable From Random

arXiv:2609.00576v1 Announce Type: new Abstract: Item-sensitivity, defined as whether a model's choice depends on the specific input rather than on its own output prior, is widely reported as evidence

Do General NLP Embeddings Capture Ontological Reasoning?

model-releasesarxiv-cs-ai

arXiv:2609.00177v1 Announce Type: cross Abstract: General-purpose NLP embedding models perform well on linguistic tasks, but their ability to capture symbolic ontological structure remains unclear. We

EM^2Mem: Event-Centric Multimodal Memory for Large Language Models

arXiv:2609.00551v1 Announce Type: cross Abstract: Multimodal memory offers a scalable interface for long-video question answering, but existing methods often retrieve captions, frames, transcripts, su

Evaluating Second-Order Bias of LLMs Through Epistemic Entitlement

arXiv:2606.17506v2 Announce Type: replace Abstract: Evaluations of social bias in LLMs largely focus on whether models generate or imply biased content. However, as LLMs are increasingly used as judge

safetyarxiv-cs-cl

From Detection to Refusal: Safer LLMs via Circuit-Guided Weight Scaling

arXiv:2609.00051v1 Announce Type: cross Abstract: Despite extensive alignment efforts, Large Language Models (LLMs) remain vulnerable to generating unsafe content under adversarial prompting, yet the

Frozen Cores Need Task Signal: Fisher-Whitened Cross-Covariance for Low-Resource LLM Adaptation

model-releasesarxiv-cs-lg

arXiv:2609.00762v1 Announce Type: new Abstract: Parameter-efficient fine-tuning is usually framed as a question of how many parameters to update. Under a severe trainable-state budget, however, where

Local Ai

Instance-Guided Report Anchoring for Text-Free 3D Abnormality Segmentation in Chest CT

arXiv:2609.00447v1 Announce Type: new Abstract: Accurate 3D abnormality segmentation in chest CT requires dense spatial supervision, but obtaining expert voxel-level labels is costly. Radiology report

local-aiarxiv-cs-cv

Knowing When to Stop: Adaptive Action Chunking via Internal Cross-Attention Dynamics in VLAs

arXiv:2609.00908v1 Announce Type: new Abstract: Action chunking is a standard execution strategy in modern Vision-Language-Action (VLA) frameworks, but fixed execution horizons impose a trade-off betw

safetyarxiv-cs-ro

LatentPress: Context Compression Beyond Text and Vision

arXiv:2609.01507v1 Announce Type: cross Abstract: Compressed context is usually carried as human-readable text or as rendered images that must be decoded, even when its consumer is a language model. W

Agents

Learning What to Retain: Gated-Memory Routing for Efficient Collaboration in Multi-Agent LLM Systems

arXiv:2609.00237v1 Announce Type: new Abstract: Large language model (LLM)-based multi-agent systems tackle complex reasoning by orchestrating how multiple agents are configured and how they collabora

agentsarxiv-cs-ai

Mac Heads: Is there any point to MLX in September 2026?

model-releasesr-localllama

This may be somewhat specific to Qwen3.8 27b and the Apple M5 series, perhaps, but enough of us are running this combo that it's worth tossing out there. GGUF models with MTP have been the fastest way

Tutorials

Manifold-Aware General Coded Computing for Straggler-Resilient Distributed Computing

arXiv:2609.00552v1 Announce Type: new Abstract: Existing coded-computing designs do not explicitly exploit the intrinsic structure of the input data. In communication systems, statistical structure an

tutorialsarxiv-cs-lg

Modeling Information Blackouts in Missing Not-At-Random Time Series Data

model-releasesarxiv-cs-lg

arXiv:2601.01480v3 Announce Type: replace-cross Abstract: Traffic forecasting systems rely on fixed sensor networks that frequently exhibit contiguous blackouts. Such outages are usually treated as ig

MUGEN: Generating Unlearnable Graph Examples for Multiple Learning Tasks

arXiv:2609.00696v1 Announce Type: new Abstract: Graph data across diverse domains can expose valuable relational information to unauthorized representation learning, creating a pressing need for prote

researcharxiv-cs-lg

Applications

NeuroPriv: Adversarial Representation Learning for Privacy in Wearable EEG Systems

arXiv:2609.00390v1 Announce Type: cross Abstract: Wearable EEG systems may expose sensitive information beyond their intended health function, creating substantial risks to neuroprivacy. In this work,

applicationsarxiv-cs-lg

Applications

No Pixel Left Behind: Filling Gaps in Anime Colorization

arXiv:2609.00800v1 Announce Type: cross Abstract: Animation production workflows often involve digital colorization of line art, where small unpainted regions ('gaps') frequently occur and remain an u

applicationsarxiv-cs-cv

Physically Plausible Video Generation via Visual-Semantic Chain-of-Events Conditioning

arXiv:2609.00656v1 Announce Type: new Abstract: Physically Plausible Video Generation (PPVG) seeks to synthesize videos consistent with physical principles, yet remains challenging due to underspecifi

researcharxiv-cs-cv

PopPert: Population-level Joint-Distribution Modeling for Single-Cell Perturbation Prediction

arXiv:2609.01357v1 Announce Type: cross Abstract: Predicting transcriptional responses to specific perturbations is critical for understanding cellular regulatory mechanisms and accelerating drug disc

Predicting Program Exit Code with LLMs and Programming Language Semantics

arXiv:2609.00579v1 Announce Type: cross Abstract: Large language models (LLMs) have shown proficiency in various software engineering tasks, such as code generation and translation. However, a key lim

Prompt-Robust Language Models: Which Training Strategies Work?

arXiv:2609.01217v1 Announce Type: new Abstract: Despite their strong performance, large language models remain highly sensitive to prompt formulation. Prior work addresses this through refined data co

Quit While You're Ahead: Quit for Efficient Candidate Generation in Machine Translation Reranking

arXiv:2609.00588v1 Announce Type: new Abstract: Reranking methods, such as Minimum Bayes Risk (MBR) decoding and Quality Estimation (QE) reranking, are widely used in modern neural machine translation

researcharxiv-cs-cl

Hardware

Reconstruct! Don't Encode: Self-Supervised Representation Reconstruction Loss for High-Intelligibility and Low-Latency Streaming Neural Audio Codec

arXiv:2603.05887v2 Announce Type: replace-cross Abstract: Neural audio codecs optimized for mel-spectrogram reconstruction often fail to preserve intelligibility. While semantic encoder distillation i

hardwarearxiv-cs-ai

SlideBank: A Persistent Hierarchical Evidence Bank for Consistent Whole-Slide Reasoning

arXiv:2609.00342v1 Announce Type: new Abstract: Whole-slide images (WSIs) are challenging for vision-language reasoning because diagnostically relevant morphology is sparse, heterogeneous, and distrib

SMELT: Scaling Laws for Compute-Matched MoE Looped Transformers

arXiv:2609.01343v1 Announce Type: new Abstract: Looped Transformers increase effective depth by iterating a shared block of layers, but most evaluations compare at fixed model size, conflating archite

safetyarxiv-cs-lg

StreamScout: Learning When to Look Deeper for Streaming Video Understanding

arXiv:2609.00291v1 Announce Type: new Abstract: Streaming video understanding requires answering questions that arrive at arbitrary moments over an unbounded video stream. Existing systems primarily f

safetyarxiv-cs-cv

Applications

Teaching Vision-Language Models to Use the Scale They Are Given: Label-Free Equivariance Training for Metric Physical Reasoning

arXiv:2609.00658v1 Announce Type: new Abstract: Metric questions about video require vision-language models to use supplied real-world references to convert visual measurements into physical units. Ye

applicationsarxiv-cs-cv

Towards a Belief-Based World Model for LLM Agents

model-releasesarxiv-cs-ai

arXiv:2609.00455v1 Announce Type: new Abstract: Large language models (LLMs) are being used as policies for autonomous decision-making and planning in many domains. Despite their strong reasoning capa

Triple-Bottom-Line Sustainability of Language Models for Edge AI: A Comparison Between SLMs and Quantized LLMs

model-releasesarxiv-cs-ai

arXiv:2609.00665v1 Announce Type: new Abstract: Edge-AI model selection is commonly driven by one isolated metric - accuracy, latency, memory, energy, or safety, even though a deployable language mode

Validity-Aware Jailbreak Evaluation for Large Language Models

arXiv:2609.00498v1 Announce Type: new Abstract: Jailbreak robustness has become central to large language model (LLM) safety evaluation, yet prevailing methodologies rely primarily on refusal behavior

Local Ai

Vision-Based Leader-Follower Formation Control for Cooperative UAVs in GPS-Degraded Environments

arXiv:2609.01420v1 Announce Type: new Abstract: Cooperation in multi-UAV systems requires reliable relative perception so that follower vehicles can maintain formation and continue their mission safel

local-aiarxiv-cs-ro

We benchmark Qwen3.6-35B-A3B on an M5 Max MacBook Pro. Across ten prompt lengths and ten decode contexts, Lily was faster than MLX-LM, avera…

model-releasesperplexity--x

We benchmark Qwen3.6-35B-A3B on an M5 Max MacBook Pro. Across ten prompt lengths and ten decode contexts, Lily was faster than MLX-LM, averaging 1.23× higher prefill throughput and 1.35× higher decode

When Compression Helps and When It Hurts: Condition-Aware Analysis of Chain-of-Thought Distillation

arXiv:2606.21704v2 Announce Type: replace Abstract: Chain-of-Thought (CoT) distillation transfers multi-step reasoning from large reasoning models to smaller students, but verbose teacher traces infla

researcharxiv-cs-cl

When Safety Routing Breaks: Understanding Alignment Fragility under Benign Fine-Tuning

arXiv:2609.01455v1 Announce Type: cross Abstract: Benign fine-tuning severely weakens the safety alignment of large language models (LLMs), so we study why refusal behavior is so fragile. While prior

WiSDoM: Wireless Sparse Decision Transformer with Mixture-of-Experts for Multi-Task Mobile Network Optimization

arXiv:2609.00284v1 Announce Type: cross Abstract: Emerging 6G wireless networks are expected to operate across diverse deployment scenarios, where variations in network topology, user mobility, traffi

Workload Identification with Physical Side Channels for AI Governance

arXiv:2609.00309v1 Announce Type: cross Abstract: AI compute verification is one of the first tangible and tractable points for international policy aimed at AI governance. Determining whether frontie

Accelerating Unified Multimodal Models with Core-Expansion Routing and Unified Computation Scheduling

arXiv:2608.29291v1 Announce Type: new Abstract: Unified multimodal models jointly support understanding and generation, but incur substantial redundant computation across tokens, layers, and generatio

AcrossWAM1.0:A Modular Latent World-Action Stack for Compact Robot Policies

arXiv:2608.29937v1 Announce Type: new Abstract: Latent world-action models avoid rendering future pixels by predicting an action-relevant visual subgoal in feature space. LaWAM established this formul

Adaptive Multi-Branching for Shallow Decision Tree Induction

arXiv:2608.29262v1 Announce Type: cross Abstract: Decision trees are attractive for tabular prediction tasks because each prediction follows an interpretable sequence of feature-threshold tests. Under

AlgoWorlds: Benchmarking Tool Use for Global Optimization in Algorithmic Worlds

model-releasesarxiv-cs-cl

arXiv:2608.29397v1 Announce Type: new Abstract: Tool-use benchmarks generally evaluate whether an agent completes a workflow using appropriate tools and valid arguments. However, feasibility alone is

Applications of Risk Science to AI Fairness Evaluation: Principles, Challenges, and Best Practices

arXiv:2608.29478v1 Announce Type: cross Abstract: Scholarly work which aims to describe potential societal impacts (e.g., risks) of proliferating technology (especially related to artificial intellige

Are Online Skill and Memory Modules Always Worth Their Tokens? A Budget-Constrained Study of Web Agents

model-releasesarxiv-cs-cl

arXiv:2606.15017v2 Announce Type: replace Abstract: Online web agents often augment a base actor with memory, workflow, or skill modules. These modules can improve performance, but they also consume t

Artifact Reduction in Undersampled 3D Cone-Beam CTs using a Hybrid 2D-3D CNN Framework

arXiv:2602.08727v2 Announce Type: replace-cross Abstract: Undersampled CT volumes minimize acquisition time and radiation exposure but introduce artifacts degrading image quality and diagnostic utilit

Auditing and Mitigating Privacy Leakage in Cloud-Edge Collaborative Decoding

arXiv:2608.29111v1 Announce Type: cross Abstract: Applications such as personalized assistance and proprietary document analysis require large language models (LLMs) to generate outputs from private d