Search: “llm”

Contextualized Counterspeech Can Be More Persuasive Than Generic Counterspeech

arXiv:2607.26236v1 Announce Type: cross Abstract: AI-generated counterspeech offers a scalable and effective strategy to mitigate online toxicity by promoting more constructive dialogue. Yet, existing

researcharxiv-cs-ai

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Contrastive Reinforced Policy Optimization via Privileged Self-Distillation

DGX agent

arXiv:2607.28026v1 Announce Type: new Abstract: Recent advances in post-training Large Language Models (LLMs) increasingly rely on Reinforcement Learning with Verifiable Rewards (RLVR) or On-Policy Se

safetyarxiv-cs-lg

31 Jul 2026

datasette-agent 0.4a0

Release: datasette-agent 0.4a0 New await context.browser_task() mechanism allowing agent tools to run code directly in the user's browser. #33 This is an exciting new capability: it makes it easy for

agentssimon-willison

Decision-oriented joint optimization of evidence fusion based on event-conditioned credibility

arXiv:2504.04128v3 Announce Type: replace Abstract: In decision-level fusion tasks involving heterogeneous sources with unequal precision and potential anomalies, evidence deviating from the majority

researcharxiv-cs-ai

Distributions In, Distributions Out: The Case for Soft-Label Training

arXiv:2511.14117v2 Announce Type: replace Abstract: Supervised classifiers output a distribution over classes but are typically trained against a single label obtained by collapsing multiple annotator

researcharxiv-cs-lg

Divergence Decoding: Training-Free Capability Fusion

arXiv:2607.27248v1 Announce Type: cross Abstract: While large language models excel in reasoning, these generalists often lack knowledge for specialized scientific domains. Conversely, domain models~(

Tutorials

Dynamically Scaled Activation Steering

arXiv:2512.03661v2 Announce Type: replace Abstract: Activation steering has emerged as a powerful method for guiding the behavior of generative models towards desired outcomes such as toxicity mitigat

tutorialsarxiv-cs-lg

EMBL AI Librarian: Life-Sciences Knowledge Layer for AI Agents

arXiv:2607.28229v1 Announce Type: new Abstract: The web is increasingly accessed by AI agents rather than humans. Every agent needs knowledge, especially in the life-sciences, where agentic pipelines

Energy-Driven Adaptive Visual Token Pruning for Efficient Vision-Language Models

arXiv:2603.05950v2 Announce Type: replace Abstract: Visual token reduction is critical for accelerating Vision-Language Models (VLMs), since visual inputs are represented as token sequences that intro

researcharxiv-cs-cv

Exact Action Values Are Not Enough: Rollout-Verified Reinforcement Fine-Tuning of a Reasoning Model for Multi-Zone VAV Control

arXiv:2607.27914v1 Announce Type: new Abstract: Multi-zone variable-air-volume control must balance thermal comfort, indoor air quality, and electricity use across several continuous actuators. Model

Hardware

FinCacheServe: Dependency-Consistent Answer Reuse for Cost-Efficient RAG Serving over Mutable Enterprise Documents

arXiv:2607.26076v1 Announce Type: cross Abstract: Retrieval-augmented generation services over mutable enterprise documents repeatedly execute semantically equivalent analysis requests. Answer reuse c

hardwarearxiv-cs-ai

Flux-OPD: On-Policy Distillation with Evolving Contexts

arXiv:2607.28022v1 Announce Type: new Abstract: Large language model training in open-ended domains lacks verifiable rewards, making task preferences difficult to formalize as effective supervision. C

safetyarxiv-cs-lg

From Single- to Cross-Document: Benchmarking Multi-Granularity Event Analysis of Large Language Models

arXiv:2607.27654v1 Announce Type: new Abstract: Event analysis is an essential and fundamental direction of information extraction, involving various event-centric tasks at different granularity of do

Group-Reflective Self-Distillation for Agentic Reinforcement Learning

arXiv:2607.28076v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) is effective for training large language model agents. However, terminal rewards provide only co

safetyarxiv-cs-lg

Harness-G: A Graph-Structured Harness for Search Agents

arXiv:2607.27652v1 Announce Type: new Abstract: Reinforcement learning (RL) search agents commonly model retrieval as free-form natural-language query generation and optimize multi-turn interactions u

safetyarxiv-cs-cl

Language Diversity: Evaluating Language Usage and AI Performance on African Languages in Digital Spaces

arXiv:2512.01557v3 Announce Type: replace Abstract: This study examines the digital representation of African languages and the challenges this presents for current language detection tools. We evalua

Memory Decoder at Scale: A Pretrained, Parametric Long-Term Memory

arXiv:2607.27919v1 Announce Type: new Abstract: Decoder-only language models entangle long-term memory and reasoning in a single parameter set, making it difficult to scale memory capacity independent

MMOOC: A Comprehensive Benchmark for Out-of-Context Evaluation in Multimodal Large Language Models

model-releasesarxiv-cs-cv

arXiv:2607.27637v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) have achieved strong performance on a wide range of vision-language tasks, but often fail under imperfect or sh

Hardware

ObjectStream: Latent Objects as Memory Anchors for Streaming Video Understanding

arXiv:2607.28312v1 Announce Type: new Abstract: Streaming video understanding requires models to continuously retain useful visual evidence before future questions are known. Existing approaches prima

hardwarearxiv-cs-cv

ORCA-bench: How Ready Are Language Model Agents for Oncall?

arXiv:2607.28545v1 Announce Type: new Abstract: Large language models can write, patch, and search code, but oncall root cause analysis (RCA) demands something different: reasoning over noisy metrics,

PlatformBid: An Auto-Bidding Benchmark from a Unified Advertising Platform's Perspective

arXiv:2607.27265v1 Announce Type: new Abstract: Real-time bidding is central to computational advertising, comprising three elements: Supply Side Platform (SSP) selling ad impressions, Demand Side Pla

Policy Gradient Steering: Interventions from Behavioral Objectives

arXiv:2607.27574v1 Announce Type: new Abstract: Activation steering has emerged in large language models as a lightweight alternative for dynamically changing a model's behavior at inference time. How

safetyarxiv-cs-lg

Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents

model-releasesarxiv-cs-cv

arXiv:2607.28227v1 Announce Type: cross Abstract: GUI agents have the potential to become a general purpose executor over existing digital devices. To advance them toward real-world use, we envision a

Recall Before You Rank: Similarity-Guided Top-K Reuse for Efficient Long-Context Attention

arXiv:2607.27692v1 Announce Type: new Abstract: Top-K sparse attention reduces the cost of Softmax and value aggregation by attending to only a small subset of key--value (KV) entries. However, identi

Tutorials

Relational Scene Graphs for Object Grounding of Natural Language Commands

arXiv:2602.04635v2 Announce Type: replace Abstract: Robots are finding wider adoption in human environments, increasing the need for natural human-robot interaction. However, understanding a natural l

tutorialsarxiv-cs-ro

RepBench: Compiling Benchmarks into Capability Representations for Large Language Models

arXiv:2607.28008v1 Announce Type: new Abstract: Representation engineering reads and steers capability directions in large language models, yet methods are typically evaluated on paper-specific synthe

Tutorials

Same Graph Cross-Task Transfer in GNNs: Protocols and Predictors

arXiv:2607.28525v1 Announce Type: new Abstract: Many real-world graphs support multiple predictive tasks over the same underlying structure, creating an opportunity to reuse supervision across node cl

tutorialsarxiv-cs-lg

SciDataSailor: Deep Scientific Data Exploring

arXiv:2607.28098v1 Announce Type: cross Abstract: Scientific datasets are commonly organized as hierarchical repositories containing heterogeneous and interdependent files, making their inspection, in

agentsarxiv-cs-cl

SciSchema.org: A Multidisciplinary Collection of Schemas for Structured Scientific Process Descriptions

arXiv:2607.27955v1 Announce Type: cross Abstract: Scientific processes are often described in heterogeneous article discourse, with details needed for comparison, reproducibility, reuse, and automatio

Semantic-Aligned Structural Abstraction for Multimodal Sentiment Analysis

arXiv:2607.27790v1 Announce Type: new Abstract: Multimodal Sentiment Analysis (MSA) aims to interpret complex human emotions by integrating natural language with non-verbal modalities. Non-verbal moda

smevals - a small eval suite for evaluating models, prompts, and harnesses

model-releasessimon-willison

smevals - a small eval suite for evaluating models, prompts, and harnesses I've been working with Jesse Vincent's Prime Radiant applied AI research lab building out this evals framework to help answer

Strategy, Not Payoffs: A Behavioural Embedding of Normal-Form Games

arXiv:2607.27536v1 Announce Type: cross Abstract: Learning a strategic task changes more than what is directly taught: fine-tuning on one game can either enhance or degrade an agent's ability to reaso

agentsarxiv-cs-lg

TCA-SIR: Learning Target-Conditioned Abstractions for Scientific Inspiration Retrieval

arXiv:2607.28498v1 Announce Type: cross Abstract: Scientific hypothesis generation for AI for Science typically involves Scientific Inspiration Retrieval (SIR) followed by hypothesis composition. Exis

ThreatForest: Multi-Agent Attack Tree Generation with Pluggable TTP Framework Mapping

arXiv:2607.27528v1 Announce Type: cross Abstract: Threat modeling is essential for secure software development, yet manual analysis of cloud-native architectures is slow and demands scarce security ex

agentsarxiv-cs-cl

Towards Structurally Explainable Machine-Generated Text Detection: A Graph-Perspective Framework

arXiv:2505.12507v2 Announce Type: replace Abstract: Despite the success of machine-generated text detectors, the black-box nature remains a critical limitation. Traditional explainability methods rely

Hardware

Understanding Is Done Early: A Depth Division of Labor in Large Language Models and Its Use for Unbounded-Context Memory

arXiv:2607.28263v1 Announce Type: new Abstract: Transformer depth is not used uniformly: lower and middle layers build semantic representations, while upper layers increasingly specialize them for pre

hardwarearxiv-cs-cl

VETO: Towards Protecting Images From Frontier AI Editing

arXiv:2607.27292v1 Announce Type: new Abstract: The rise of powerful, accessible image-editing models such as FLUX.2 has brought high-fidelity editing within broad reach. Their capabilities now extend

researcharxiv-cs-cv

What Does It Take to Detect an AI Agent? Minimal Feature Sets for Behavioral Detection under Browser Automation

model-releasesarxiv-cs-ai

arXiv:2607.26935v1 Announce Type: new Abstract: Bot detectors deployed at scale treat traffic as binary: human or bot. This assumption breaks when AI agents browse the web through browser automation,

WhisperRec: Latent Reasoning for Efficient Foundation Recommendation Models

model-releasesarxiv-cs-ai

arXiv:2607.26621v2 Announce Type: cross Abstract: Large language models (LLMs) have demonstrated strong reasoning capabilities, motivating their adoption as backbones for foundation recommendation mod

Local Ai

2 images + 1 prompt > expected output

Hi, I'm trying to replicate a thing locally, that I can do on ChatGPT. What I want is to give a local AI two reference images (a face and a background item) and a prompt about the composition of the p

local-air-ollama

Local Ai

2× Radeon R9700 for Local AI Was Choosing AMD Instead of NVIDIA a Mistake Without CUDA?

Hello together I decided to go with 2× Radeon AI PRO R9700 GPUs (64 GB total VRAM) for my local AI server. However, I keep reading that AMD/ROCm is still not as mature as NVIDIA/CUDA when it comes to

local-air-localllama

Advancing the price-performance frontier with GPT‑5.6

model-releasessimon-willison

Advancing the price-performance frontier with GPT‑5.6 Huge price drop from OpenAI today: GPT-5.6 Terra got a 20% reduction, and GPT-5.6 Luna got a massive 80% drop. OpenAI credit 5.6 Sol with enabling

ARC-Encoder: learning compressed text representations for large language models

arXiv:2510.20535v2 Announce Type: replace Abstract: Recent techniques such as retrieval-augmented generation or chain-of-thought reasoning have led to longer contexts and increased inference costs. Co

Decoupled Visual Processing: Efficient Multimodal Adaptation via Modality-Specific Transformer Substitution

model-releasesarxiv-cs-cv

arXiv:2607.26596v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) have demonstrated remarkable capabilities by integrating visual and textual understanding within a unified tran

Deep Expert Injection for Anchoring Retinal VLMs with Domain-Specific Knowledge

arXiv:2603.07131v4 Announce Type: replace Abstract: Large Vision Language Models (LVLMs) show immense potential for automated ophthalmic diagnosis. However, their clinical deployment is severely hinde

researcharxiv-cs-cv

DuplexGen: Adaptive Synthesis of Human-AI Turn-Taking Dialogues

arXiv:2607.26178v1 Announce Type: new Abstract: Turn-taking is a central component of full-duplex interaction. Which turn-taking behaviors are appropriate varies with the scenario, yet current models

EvoPINN: Agentic Discovery of Executable Algorithms for Physics-Informed Neural Networks

arXiv:2607.26490v1 Announce Type: cross Abstract: Physics-informed neural networks (PINNs) have emerged as a powerful paradigm for solving partial differential equations (PDEs), yet their performance

FedWeave: Rethinking the Unit of Specialization in Heterogeneous Federated MoE-LoRA

arXiv:2607.26618v1 Announce Type: cross Abstract: Federated PEFT enables LLMs to collaboratively adapt to decentralized private data without sharing raw examples. However, task heterogeneity across cl