Search: “reasoning”

1/2 Thanks Gavin for an especially thoughtful exchange. I don't usually spend much time on social media but I wanted to engage here because …

1/2 Thanks Gavin for an especially thoughtful exchange. I don't usually spend much time on social media but I wanted to engage here because it really brings out the heart of an important conversation.

safetydario-amodei--x

15 Aug 2026

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

club-5060ti refresh: tested RTX 5060 Ti presets, a proper high-context harness, and Qwen3.8 27B

DGX agent

Quick update on the RTX 5060 Ti local LLM repo. It has changed quite a bit since my previous posts. The project started as a collection of practical notes and benchmark results. That was useful, but a

model-releasesr-localllama

15 Aug 2026

SOTA Apple Silicon Inference (August 15, 2026)

model-releasesr-localllama

This is a HANDWRITTEN post. I spent way too much time trying to get fast inference on Apple Silicon. This post is for people who want to know what's the latest on running local models on their mac, an

15 Aug 2026

Agent Behavioral Contracts II: Certifying Compositional Reliability Without Assuming Independence

arXiv:2608.12895v1 Announce Type: new Abstract: Compositional reliability bounds for multi-agent systems multiply component reliabilities, a step licensed by a conditional-independence assumption that

AirForesight: Current-to-Future Spatial Map Imagination with Cross-Space Planning Consistency for UAV-VLN

arXiv:2608.12835v1 Announce Type: new Abstract: Unmanned Aerial Vehicle Vision-Language Navigation (UAV-VLN) requires agents to follow language instructions, infer spatial structure from sparse multi-

researcharxiv-cs-ro

arepsilon-MemEvo: Adaptive Cross-Task Memory Transfer for LLM Program Evolution

arXiv:2608.12522v1 Announce Type: new Abstract: LLM-based program evolution systems such as FunSearch and AlphaEvolve have shown strong ability to discover novel algorithms, but typically optimize eac

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents

arXiv:2608.12977v1 Announce Type: cross Abstract: The expanding operational capabilities of large language model (LLM) agents introduce sophisticated security threats. Runtime defenses have emerged as

Beyond Outcome Rewards: Step-Level Self-Distilled Policy Optimization for Deep Search Agents

arXiv:2608.12764v1 Announce Type: cross Abstract: Deep search agents operate over trajectories spanning dozens of steps, yet standard reinforcement learning provides only a single outcome reward per t

BrainWAM: Action-Space Coordination of Semantic Priors and Predictive Dynamics for Autonomous Driving

arXiv:2608.12854v1 Announce Type: cross Abstract: Autonomous driving requires planning under both semantic constraints and predictive dynamics. Existing end-to-end driving approaches, however, typical

Tutorials

Class Geometry as Supervision for Sample-Efficient Open-World Detection

arXiv:2608.12698v1 Announce Type: new Abstract: Open-world object detection requires models to recognize known categories, reject unfamiliar objects, and incorporate new classes over time. This is esp

tutorialsarxiv-cs-cv

CROP: Task Relevance via Counterfactuals for Selective On-Policy Distillation

arXiv:2608.13387v1 Announce Type: new Abstract: On-policy distillation (OPD) supervises a student language model on trajectories sampled from its current policy, but assigns equal credit to response t

safetyarxiv-cs-cl

Designing AI Pipelines for Decision-Ready ITSM Intelligence

arXiv:2608.12670v1 Announce Type: new Abstract: IT service management (ITSM) systems accumulate large volumes of heterogeneous ticket data that are difficult for sales and executive stakeholders to co

Diagnostic Foundation for Evaluating LLMs' Research Integrity as Co-Scientists

arXiv:2608.12345v1 Announce Type: new Abstract: Language models are increasingly deployed as co-scientists, yet their ability to uphold research integrity under institutional pressure remains unmeasur

DiffGRM: Diffusion-based Generative Recommendation Model

arXiv:2510.21805v2 Announce Type: replace-cross Abstract: Generative recommendation (GR) is an emerging paradigm that represents each item via a tokenizer as an n-digit semantic ID (SID) and predicts

DMDIntel: Interpreting Large Language Models via Dynamic Mode Decomposition

arXiv:2608.13048v1 Announce Type: new Abstract: In this work, we introduce DMDIntel which uses dynamic mode decomposition (DMD) to make the predictions made by LLMs in a classification task interpreta

Applications

EgoPHI: Estimating Contact and Force from Egocentric Vision

arXiv:2608.13014v1 Announce Type: new Abstract: Understanding hand-object interaction from egocentric vision is essential for modeling how people physically engage with the surrounding world. Yet reas

applicationsarxiv-cs-cv

Enhancing In-Hospital Mortality Prediction Using Multi-Representational Learning with LLM-Generated Expert Summaries

arXiv:2411.16818v2 Announce Type: replace-cross Abstract: To evaluate a multi-representational framework in which large language model (LLM)-generated expert summaries of intensive care unit (ICU) not

Foam-Agent: A Large Language Model-Based Multi-Agent Framework for Automating Computational Fluid Dynamics Workflows

arXiv:2505.04997v3 Announce Type: replace Abstract: Computational fluid dynamics (CFD) has been the main workhorse of computational physics, yet its steep learning curve and fragmented, multi-stage wo

From Refuse to Richness: Rubric Rewards for Long-Form Hallucination Reinforcement Learning

arXiv:2608.12337v1 Announce Type: new Abstract: Rewards that penalize unsupported claims can improve grounding in long-form generation, but they can also teach models to answer less. We study this ref

Geometric and Behavioral Stratification in Transformer Residual Streams

arXiv:2608.12447v1 Announce Type: cross Abstract: Trained transformer models develop privileged bases: coordinate axes whose statistics differ from the rest of the residual stream. But what kind of di

Gradual Code-Switching as Inference-Time Cross-Lingual Representational Alignment for LLMs

arXiv:2510.05678v2 Announce Type: replace-cross Abstract: While large language models (LLMs) have achieved notable progress in multilingual settings, their performance remains uneven across languages

Heterogeneous Vision-Language Ensemble with Disagreement-Aware Reranking for Text-Based Person Anomaly Retrieval

arXiv:2608.12843v1 Announce Type: cross Abstract: Text-based person anomaly retrieval aims to retrieve pedestrians exhibiting anomalous behaviors from a large image gallery using natural language desc

Intensional Anaphora

arXiv:2608.12598v1 Announce Type: new Abstract: Intensional operators are often treated as quantifiers over possible worlds, parallel to the treatment of determiners as quantifiers over individuals. Y

Is this Citation on Point?

model-releasesarxiv-cs-cl

arXiv:2608.12571v1 Announce Type: cross Abstract: In 2023, a New York judge sanctioned two attorneys in Mata v. Avianca for filing a brief with hallucinated citations generated by ChatGPT. Such failur

Tutorials

Jointly Predicting Courses and Grades Using a Transformer-Based Model

arXiv:2608.13409v1 Announce Type: new Abstract: Existing predictive models in learning analytics often treat student academic history as a simple sequence, overlooking the concurrent nature of courses

tutorialsarxiv-cs-ai

Local Ai

LLM-Assisted Dynamic Threat Analysis for Attacker-Reachable Software Weaknesses in Autonomous Vehicles

arXiv:2608.13450v1 Announce Type: cross Abstract: Autonomous vehicles depend on large safety-critical software stacks, where weaknesses reachable from adversarial inputs may affect steering, braking,

local-aiarxiv-cs-lg

MOSAIC: Unveiling the Moral, Social and Individual Dimensions of Large Language Models

arXiv:2603.00048v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly deployed in sensitive applications including psychological support, healthcare, and high-stakes

Not All Nudges Land: Behavioral Controllability and Elaboration Quality in AI-Supported Journaling

arXiv:2608.12582v1 Announce Type: cross Abstract: AI journaling tools can tailor prompts to a person's own sensed behavior, but it is unclear which behaviors respond to them. We analyzed 369 journal e

OmniScientist: An Omni-Modal Omni-Discipline AI Scientist

arXiv:2608.13558v1 Announce Type: new Abstract: Recent advances in foundation models have enabled AI scientists to automate increasingly complete research workflows, from hypothesis generation and cod

PatientAct: Theory-Grounded Mental Health Client Simulation

arXiv:2608.12750v1 Announce Type: cross Abstract: LLM-based simulated clients are increasingly used to train novice counselors, evaluate LLM therapists, and generate synthetic data. However, current s

Polish Medical Visual Question Answering: Vision-Language Models Underutilize Visual Evidence

arXiv:2608.12928v1 Announce Type: new Abstract: We introduce a Polish-language medical visual question answering (VQA) benchmark, built from Polish Board Certification Examination questions for licens

RetroCraft - Qwen 3.8 27B Q8, one shot with exact performance data on dual 3090s.

model-releasesr-localllama

The demo video above was created with a single prompt using Qwen 3.8 27B Q8_K_XL. Prompt: create me a single page, html/js/css GeoCities, Laser Wave, 80's retro, minecraft clone named RetroCraft. Embe

SDS-LoRA: Overcoming Anisotropic Gradient Scaling in Low-Rank Adaptation

arXiv:2606.16454v2 Announce Type: replace-cross Abstract: Low-Rank Adaptation (LoRA) enables efficient adaptation of large pretrained models to downstream tasks by parameterizing weight updates with l

SkillEvo: Self-Renewing Evolution Gradients from Multi-Turn Interaction Feedback

arXiv:2608.13120v1 Announce Type: new Abstract: Agent Skills are today either hand-authored or produced in a single LLM generation pass, and consequently possess no closed loop through which they migh

SpatialVAM:Spatial-Aware Multi-View Video Diffusion as a Data-Efficient Robot Policy

arXiv:2604.03181v2 Announce Type: replace-cross Abstract: Robotic manipulation requires understanding both the 3D spatial structure of the environment and its temporal evolution, yet most existing pol

safetyarxiv-cs-cv

StreamTTT: Reconciling Real-Time Perception and Long-Term Memory in Streaming VLMs

model-releasesarxiv-cs-cv

arXiv:2608.13416v1 Announce Type: new Abstract: Humans effortlessly perceive the present while remembering the past, yet streaming VLMs often trade off real-time perception against long-term memory. P

The Objective Is the Bottleneck: Latent World Models Encode What Their Planners Cannot Use

arXiv:2608.12959v1 Announce Type: cross Abstract: Latent world models are judged by how well they predict, so when planning fails at long horizons the natural reading is that the predictor degrades. O

The Role of Natural Language Understanding in Multimodal Video-Based Dengue Diagnosis

arXiv:2608.12677v1 Announce Type: new Abstract: Detecting infection-related behavioral changes in mosquitoes from video data is challenging because mosquitoes are small, move rapidly and irregularly,

TopoIntent: Compiling Security Intent into Executable, Compliance-Checked Network Topologies

arXiv:2608.13389v1 Announce Type: new Abstract: Enterprise security topology design requires translating business intent, regulatory requirements, and risk assumptions into zones, boundary devices, in

Applications

Towards Physics-Faithful Generation of Scientific Diagrams

arXiv:2608.13112v1 Announce Type: new Abstract: Text-to-image generation has reached photorealistic quality, yet state-of-the-art systems remain unreliable at producing scientific diagrams, whose valu

applicationsarxiv-cs-cv

Applications

Towards Socially Compliant Navigation in Deep Reinforcement Learning via Proxemics-Based Reward Modeling

arXiv:2608.12917v1 Announce Type: new Abstract: Developing effective robot navigation methods in crowded environments is essential for real-world applications. Although recent deep reinforcement learn

applicationsarxiv-cs-lg

TRAPSBench: Vision-Language Models Encode but Fail to Express Epistemic Restraint

arXiv:2608.13167v1 Announce Type: cross Abstract: When visual evidence is occluded or chaotic, models should abstain. In this paper, we show that Vision-Language Models (VLMs) can internally distingui

When AI Is Your Pastor: A Benchmark for Theological Triage and Pastoral Guidance in Large Language Models

arXiv:2608.12324v1 Announce Type: cross Abstract: People increasingly ask large language models (LLMs) for counsel on questions of faith, doctrine, and pastoral care. These questions are not ordinary

When Should Multi-Round RAG Stop? Structured Stopping Judgments and Retrieval Reduction in Search-R1

arXiv:2608.13237v1 Announce Type: cross Abstract: Multi-round retrieval-augmented generation (RAG) must decide when to stop searching as evidence accumulates. Because the deployed policy is determined

safetyarxiv-cs-cl

A Reality Check of Language Models as Formalizers on Constraint Satisfaction Problems

arXiv:2505.13252v5 Announce Type: replace Abstract: Recent work shows superior performance when using large language models (LLMs) as formalizers instead of as end-to-end solvers for symbolic reasonin

Behavior and Representation in Open-Weight Large Language Models for Combinatorial Optimization: From Feature Extraction to Algorithm Selection

arXiv:2512.13374v2 Announce Type: replace Abstract: Recent advances in Large Language Models (LLMs) open new perspectives for automation in optimization, yet little is known about whether their intern

CT-DeltaBench: A Benchmark for Longitudinal 3D Medical Imaging Difference Reporting with Vision-Language Models

model-releasesarxiv-cs-cl

arXiv:2608.11534v1 Announce Type: new Abstract: In medical imaging, the clinical value of Computed Tomography (CT) lies not only in depicting current disease status, but crucially in enabling longitud

Do LLMs Take Care of Their Own? Similarity Signals Can Induce Cooperation

arXiv:2608.12125v1 Announce Type: cross Abstract: As LLM-based agents with user-instructed goals are becoming widely deployed, they increasingly encounter each other in strategic interactions, and fac