Search: “inference”

Accurate in space, unreliable in time: how LLMs represent national cultural change

arXiv:2609.01902v1 Announce Type: cross Abstract: Assessments of cultural alignment have become an important part of the development and improvement of large language models (LLMs). However, the major

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Action abstractions for amortized sampling

DGX agent

arXiv:2410.15184v2 Announce Type: replace-cross Abstract: As trajectories sampled by policies used by reinforcement learning (RL) and generative flow networks (GFlowNets) grow longer, credit assignmen

safetyarxiv-cs-ai

3 Sept 2026

ASCII Attack: Recontextualising Harmful Requests as Artistic Critique in Large Language Models

arXiv:2609.02215v1 Announce Type: new Abstract: Safety alignment trains large language models to refuse harmful requests stated plainly, but that training is applied mostly to surface form. Requests t

Banger paper from Google DeepMind and colleagues. (bookmark it) A model reads its entire KV cache on every generated token, even though it e…

Banger paper from Google DeepMind and colleagues. (bookmark it) A model reads its entire KV cache on every generated token, even though it ends up attending to a tiny slice of it. In other words, if y

model-releasesdair-ai--x

Candidate Generation and Definition-Guided Verification for Sentence-Level Depression Symptom Recognition

arXiv:2609.01833v1 Announce Type: new Abstract: Sentence-level recognition of depression symptoms is challenging because similar expressions can differ in symptom relevance, and language-model inferen

safetyarxiv-cs-cl

CHASE: Cache-Hole-Adapted Skip Exit for Looped State-Space Language Models

arXiv:2607.10110v2 Announce Type: replace Abstract: Recent work on looped language models suggests that many reasoning problems benefit from greater computational depth rather than from additional ind

CHIME: Credit-Aware Hierarchical Memory Evolution for Long-Horizon Agentic Planning

arXiv:2609.02074v1 Announce Type: new Abstract: Planning is a central capability that enables agents to decompose complex long-horizon tasks into manageable steps. Test-time search and training-based

Agents

ClaimReceipt: Verifying Evidence Sufficiency and Coverage in Agent Evaluations

arXiv:2609.01992v1 Announce Type: new Abstract: Agent evaluations face two distinct evidentiary questions: whether a reported claim is recomputable from retained evidence (sufficiency), and whether th

agentsarxiv-cs-ai

Agents

Codebook Agent: Amortized Topology Design for LLM Multi-Agent Systems

arXiv:2609.02264v1 Announce Type: new Abstract: Adapting the communication topology of an LLM multi-agent system to each query improves both accuracy and efficiency, yet current designers treat this a

agentsarxiv-cs-ai

Coverage, Not Targeting: A Structural Regime in Multi-Turn Agent Credit Assignment

arXiv:2609.02417v1 Announce Type: cross Abstract: Multi-turn agentic RL increasingly treats credit assignment as a targeting problem: given a terminal verifiable reward, per-turn methods localize cred

local-aiarxiv-cs-ai

DESA-TTA: Dynamic EMA and Source Anchoring for Test-Time Adaptation

model-releasesarxiv-cs-cv

arXiv:2609.01795v1 Announce Type: new Abstract: Vision-language object detectors (VLODs) achieve strong zero-shot performance but remain vulnerable to distribution shifts during deployment. Mean-teach

Dutch Books for Language Models

arXiv:2609.02797v1 Announce Type: cross Abstract: People increasingly use language models to support life decisions. Many such decisions involve a probabilistic forecast: How likely is a major life ev

researcharxiv-cs-ai

DynG-Diff: A State-Aware Dynamic Guidance Diffusion Framework for Probabilistic Time Series Forecasting

arXiv:2609.02068v1 Announce Type: new Abstract: Probabilistic multivariate time series (MTS) forecasting is crucial for modeling complex dynamical systems. However, existing diffusion-based methods re

local-aiarxiv-cs-lg

From Reweighting to Rewriting: Unlocking the Intervention Effects of Influential Samples in Training Data Attribution

arXiv:2609.02771v1 Announce Type: cross Abstract: Training data attribution (TDA) aims to identify training examples that shape model behavior, but its intervention value depends on both which example

Grounded, Compute-Efficient LLM Policy Agents for Energy-Poverty Equity in Physically-Constrained Peer-to-Peer Energy Markets

arXiv:2609.01918v1 Announce Type: new Abstract: Energy poverty is nearly absent from NLP-for-social-good, and the little existing work is either static retrieval/QA or relies on carbon-intensive cloud

local-aiarxiv-cs-cl

Hardware-Accelerated Instance Segmentation for Resource-Constrained Space Robotics with Criticality Analysis

arXiv:2609.02219v1 Announce Type: cross Abstract: Autonomous lunar missions require real-time per- ception under three coupled constraints: extreme low-light conditions, limited onboard compute, and r

safetyarxiv-cs-cv

Tutorials

Induction and Inquiry via Probabilistic Reasoning over Language and Code

arXiv:2609.01815v1 Announce Type: new Abstract: How humans grow and maintain abstract knowledge from the sparse, streaming noisy data of experience is a longstanding challenge in cognitive science. An

tutorialsarxiv-cs-ai

Krea2 Turbo Distill 4 step LoRA - new checkpoint (chk60K) released (fine detail now above the 8-step teacher and clean of artefacts, best prompt-adherence and teacher-faithfulness scores so far)

model-releasesr-stablediffusion

Krea 2 Turbo — 4-Step Distillation LoRA (work in progress) A LoRA for Krea 2 Turbo that reduces the minimum usable step count from 8 to 4. ⚡ Half the steps — 8 → 4, on Turbo's own deployment sigmas. ⏱

Language Models Can Control Their Own Attention

arXiv:2609.02737v1 Announce Type: cross Abstract: Language models spend most of their attention on a small fraction of context, yet they read the entire KV cache to find the few tokens that matter. If

Loom: Weaving Diagnostic Strands into Free-Text Consensus via Embedding-Space Reweighting

arXiv:2609.02649v1 Announce Type: new Abstract: Aggregating noisy, conflicting textual hypotheses into a reliable consensus is a fundamental challenge when deploying NLP systems in real-world industri

Neural Variational Cut Posteriors without Upstream Data

model-releasesarxiv-cs-lg

arXiv:2510.10268v3 Announce Type: replace-cross Abstract: In many applications, one must propagate parameter uncertainty from an earlier (upstream) analysis, available as samples, to subsequent (downs

OBJECTION! Lawyer Agents Mitigate Guilty Bias in Legal Judgment Prediction

arXiv:2609.02158v1 Announce Type: cross Abstract: Legal Judgment Prediction (LJP) models are typically trained on documents that describe facts from a prosecutorial perspective. Existing datasets furt

Online Reinforcement Learning in the Met Office Unified Model through Distributed Model-Agent Coupling

arXiv:2609.02566v1 Announce Type: new Abstract: Machine-learnt corrections can complement numerical weather prediction only if they adapt to the evolving model state while preserving dynamical consist

local-aiarxiv-cs-lg

Paper Instruments benchmarked sandbox providers for their RL rollouts, and E2B's tool execution came in up to 3x faster than every other san…

Paper Instruments benchmarked sandbox providers for their RL rollouts, and E2B's tool execution came in up to 3x faster than every other sandbox provider they tested, reducing idle GPU time and traini

safetyyohei-nakajima--x

Predict, Don't Iterate: Efficient Adaptive-Length Infilling for Diffusion Language Models

arXiv:2609.02108v1 Announce Type: cross Abstract: Diffusion language models (DLMs) have emerged as a promising alternative to the auto-regressive paradigm. With bidirectional attention and any-order g

researcharxiv-cs-ai

Tutorials

Prompting the Unknown: Understanding Response Uncertainty in Large Language Models

arXiv:2407.14845v4 Announce Type: replace-cross Abstract: Large language models (LLMs) are widely used in decision-making across diverse domains. Ensuring the generation of safe and reliable responses

tutorialsarxiv-cs-cl

Propose to Learn, Learn to Propose: Evaluability-Aware Assistance under Bounded Rationality

arXiv:2609.02242v1 Announce Type: new Abstract: AI assistants often collaborate by proposing candidate edits, plans, or designs that users evaluate before adoption. Existing assistance methods focus o

Applications

SCULPT: Training Edge Vision Models for Post-Training Quantization Readiness

arXiv:2609.01743v1 Announce Type: new Abstract: Edge vision models are difficult to deploy on resource-constrained hardware, making low-bit post-training quantization (PTQ) attractive. In practice, st

applicationsarxiv-cs-cv

SCX Router: Streaming Zero-Shot Model Selection with a Decoder-KV Classifier and a Real-World Task Ontology

arXiv:2609.02292v1 Announce Type: new Abstract: The rapid proliferation of large language models (LLMs) and the growing diversity of their applications presents a unique optimization opportunity: sele

SelfLift: Accelerating Few-Step Diffusion via Self-Recovering Resolution Transition

arXiv:2609.02036v1 Announce Type: new Abstract: Few-step diffusion models substantially compress temporal computation, making the spatial cost of each model evaluation an increasingly dominant source

local-aiarxiv-cs-cv

Industry

Sources: Mira Murati's Thinking Machines Lab is in talks to raise 1B+ at a 40B+ pre-money valuation, lower than the $50B+ valuation it sought last year (The Information)

The Information: Sources: Mira Murati's Thinking Machines Lab is in talks to raise 1B+ at a 40B+ pre-money valuation, lower than the 50B+ valuation it sought last year — Thinking Machines Lab, the AI

industrytechmeme

SR-Edit: Region-Aware Image Editing via Self-Refinement

arXiv:2609.02504v1 Announce Type: new Abstract: With the recent rapid progress in generative models, image editing has made remarkable advances, yet achieving faithful edits that precisely modify only

Training-free hack to reduce MoE reasoning tokens by 8.5% without losing accuracy but increasing Active parameters per token in MOE (Qwen 35B A4B+)

model-releasesr-localllama

Hey everyone, Wanted to share a short paper we just published exploring a simple but surprisingly effective optimization for sparse MoE reasoning models. The problem: MoE reasoning models (especially

UAV Thermal Imagery for Inert Ordnance Screening: Multi Campaign Dataset Development,Object Detection, and Practical Recommendations

arXiv:2609.01738v1 Announce Type: new Abstract: Unexploded ordnance (UXO) continues to restrict civilian access, agricultural activity, infrastructure recovery, and environmental remediation in contam

UE5M3 FP4 Block Scaling for Stable Language Model Pretraining

model-releasesarxiv-cs-lg

arXiv:2609.02846v1 Announce Type: new Abstract: Stable 4-bit floating-point (FP4) pretraining is difficult because the E2M1 payload represents only a narrow range of magnitudes. NVIDIA's Transformer E

Industry

Vals analysis: open-weight models performing multi-stage tasks, like building a web app, can have an environmental impact 10K times greater than simple queries (Bloomberg)

Bloomberg: Vals analysis: open-weight models performing multi-stage tasks, like building a web app, can have an environmental impact 10K times greater than simple queries — Benchmarker Vals AI finds t

industrytechmeme

We need a better taxonomy for what people are calling 'continual learning'

model-releasesr-localllama

Continual learning isn't some fake term but a real goal and arguably one of the most important open ones in 2026. It roughly means: ‘building systems that keep acquiring useful knowledge or skills aft

When Decodability Is Not Enough: Logical Validity Representations, Behavioral Dissociation, and Causal Tests in Language Models

arXiv:2609.02438v1 Announce Type: new Abstract: Large language models can look capable of logical reasoning, but correct or incorrect answers alone tell us little about what the model represents inter

researcharxiv-cs-cl

A Glance Is All You Need: Single-Pass Fine-Grained Image Captioning with SimLoss

arXiv:2609.00591v1 Announce Type: new Abstract: An image may be worth a thousand words, but most captioning models describe it in only a few. Modern vision-language models produce fluent high-level ca

local-aiarxiv-cs-cv

Advanced Pixel Diffusion Model with Guided Sparse Global Refinement

arXiv:2609.00798v1 Announce Type: new Abstract: Pixel-space diffusion has recently emerged as a promising direction for high-fidelity image generation by modeling images directly in the original pixel

An endless AI TV channel on a single gaming GPU — MiniMax H3, generating faster than it plays

local-air-stablediffusion

There is a video stream running on my desktop right now. It has sound, it has never repeated itself, and it will not stop. I point VLC at a local URL and it plays. One RTX 5090 does all of it — no clo

An Intelligent Decision Support System for Emotion Monitoring using Microscopic Fixational Dynamics

arXiv:2609.00846v1 Announce Type: new Abstract: The rising prevalence of psychological disorders necessitates effective emotion monitoring, yet current methods relying on facial or physiological signa

local-aiarxiv-cs-cv

Applications

Anamnesis: An Open-Source Platform for Large-Scale Backstory-Conditioned Survey Simulation

arXiv:2607.10628v2 Announce Type: replace-cross Abstract: We present Anamnesis, an interactive system for demographically controllable survey simulation using large language models. Open-source and de

applicationsarxiv-cs-ai

Any-Order GPT as Masked Diffusion Model: Decoupling Formulation and Architecture

arXiv:2506.19935v2 Announce Type: replace-cross Abstract: Efficiently scaling Large Language Models (LLMs) necessitates exploring alternatives to dominant autoregressive (AR) methods, with Masked Diff

Beyond Magnitude: Contrastive Routing for Modular Mixture-of-Experts

arXiv:2609.01100v1 Announce Type: new Abstract: In current Mixture-of-Experts architectures, routing is performed based on representations dominated by structure shared across all tokens, limiting exp

researcharxiv-cs-cl

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

CacheBridge: Efficient Cross-Model KV Cache Transfer

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

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

researcharxiv-cs-ai