Search: “agents”

DeepSeek engineer relections on RSI - burying my talent to yesterday

Note - This is translated from the actual blog link right at the bottom. A few days ago, DeepSeek v4.1 was released. It raised the ability of small models to a new level. AI is improving much faster t

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Inverting Self-Triggered Control: Adversarial Reinforcement Learning for Sparse Denial-of-Service Attacks

DGX agent

arXiv:2609.12016v1 Announce Type: new Abstract: Self-triggered reinforcement learning control (RL-STC) learns the sparsest control schedule that preserves Lyapunov-decreasing stability under a Run-Tim

safetyarxiv-cs-lg

14 Sept 2026

Not All Speech Is Intent: Adaptive Self-Correcting Inference Layer for Post-ASR False Wake-Up

arXiv:2609.12469v1 Announce Type: new Abstract: False wake-up activations remain a persistent challenge in conversational AI. Speech phonetically similar to a device's wake word can produce a syntacti

One Skill Does Not Fit All: Automatic Discovery and Taxonomy-Guided Routing of Frame-Selection Skills for Long-Video Question Answering

arXiv:2609.12517v1 Announce Type: new Abstract: Long-Video Question Answering (LVQA) requires locating decisive evidence in hour-scale videos under a limited frame budget. Most training-free methods a

PEARL: Structural Privacy-Utility Control in Human-Centric CPS via Personalized Early-Exit Deep Reinforcement Learning

arXiv:2403.05864v3 Announce Type: replace Abstract: In human-centric Cyber-Physical Systems (CPS), personalized Deep Reinforcement Learning (DRL) agents must share fine-grained control actions with cl

researcharxiv-cs-lg

Proud to report that tomorrow Sept 15th 'The AI Doc' documentary film will be streaming on Netflix. Feeling confused, anxious, or overwhelme…

Proud to report that tomorrow Sept 15th 'The AI Doc' documentary film will be streaming on Netflix. Feeling confused, anxious, or overwhelmed by the recent AI headlines, from rogue OpenAI agents carry

safetydan-hendrycks--x

READ: Learning Risk-Informed Fields for End-to-End Autonomous Driving

arXiv:2609.12371v1 Announce Type: new Abstract: Autonomous driving requires more than recognizing what is present in a scene: a planner must determine how road structure, surrounding agents, and their

safetyarxiv-cs-ro

Build around cmp 170hx for qwen flash next

Don't upvote, just looking for CPU ram options for running qwen flash next, and maybe larger models (so maybe between 128 and 256gb total ram) with an existing cmp 170hx. Use case is agentic coding so

13 Sept 2026

My experience building 64GB VRAM AI SWE assistant/agent PC

As a SWE I have ultimate belief that having a irreplaceable subscription on OpenAI and Anthropic models is actually against my beliefs (I'm not fond of becoming a digital slave and being unable to do

13 Sept 2026

Making Alternative Data Work: Context-Augmented LLMs for Financial Forecasting

arXiv:2609.11607v1 Announce Type: new Abstract: When forecasting a firm's future financial performance, alternative data - data collected from non-traditional sources such as consumer transactions, we

12 Sept 2026

What a Random Draw from the MCP Registry Contains, and What Tool-Use Benchmarks Contain Instead

arXiv:2609.10962v1 Announce Type: cross Abstract: Studies of the Model Context Protocol (MCP) server ecosystem draw their samples in ways that quietly select for servers that work: reference sets, pop

12 Sept 2026

KuaiRP Series Role-playing Models Technical Report

arXiv:2609.11127v1 Announce Type: cross Abstract: This paper introduces the complete technical solution for the KuaiRP series of role-playing models. We aim to achieve four core objectives for a dedic

Industry

OpenAI targets Wall Street bankers with a new version of ChatGPT

OpenAI Group PBC is going after Wall Street with a new version of ChatGPT that’s tailor made for financial tasks such as analysis and calculations. Predictably called ChatGPT for Financial Services, i

industrysiliconangle

Voice or Stereotype? Disentangling Acoustic and Content-Based Gender in Speech-to-Speech Models

arXiv:2609.09263v1 Announce Type: cross Abstract: Speech-to-speech (S2S) models now run inside dubbing, translation, and voice agents. Unlike text models, they hear the speaker's voice, which carries

Why Sample What You Can Enumerate? Exact Policy Optimization for Genomic Tool Selection

arXiv:2609.10221v2 Announce Type: new Abstract: Reinforcement learning over a frozen reasoner has become a common recipe for teaching a policy which external tools to invoke. We show that this recipe

Anthropic says AI now lets lone operators run state-level hacking campaigns

model-releasessiliconangle

Individual hackers are now sustaining campaigns that would have needed many skilled operators a year ago, because artificial intelligence has absorbed the labor that used to set state-backed teams apa

ASDA: Automated Skill Distillation and Adaptation for Financial Reasoning

arXiv:2603.16112v2 Announce Type: replace-cross Abstract: Adapting large language models (LLMs) to specialized financial reasoning typically requires expensive fine-tuning that produces model-locked e

Bilevel Planning with Learned Symbolic Abstractions from Interaction Data

arXiv:2603.08599v2 Announce Type: replace Abstract: Intelligent agents must reason over both continuous dynamics and discrete representations to generate effective plans in complex environments. Previ

researcharxiv-cs-ro

BRACE: Anchored Bellman-Residual Correction for Stale Critics in Asynchronous RL

arXiv:2609.09783v1 Announce Type: new Abstract: Asynchronous reinforcement learning has become the standard way to scale training for language models, but the resulting policy lag biases the critic to

safetyarxiv-cs-lg

Local Ai

Detokenization Leaks: Reconstructing Local LLM Outputs From Cache Traces

arXiv:2609.06674v1 Announce Type: cross Abstract: We present a new attack that reconstructs the text generated by locally hosted LLMs by observing CPU cache activity during detokenization. Unlike prio

local-aiarxiv-cs-ai

GANDR: Claim Auditing for Verifiable Legal Answer Generation

arXiv:2609.10293v1 Announce Type: new Abstract: In high-stakes domains such as legal practice, a language-model answer is only useful to the extent that a reader can verify each claim against the sour

Tools

GLM-5.3 is now available for training on Fireworks Dedicated Training API and Managed Training surfaces. Built for complex coding and long-h…

GLM-5.3 is now available for training on Fireworks Dedicated Training API and Managed Training surfaces. Built for complex coding and long-horizon agents and now available to train on our dedicated in

toolsfireworks-ai--x

Industry

Letter: the Senate disaster management subcommittee, led by Sen. Josh Hawley, is probing OpenAI's handling of the Hugging Face breach, calling it 'reckless' (Axios)

Axios: Letter: the Senate disaster management subcommittee, led by Sen. Josh Hawley, is probing OpenAI's handling of the Hugging Face breach, calling it “reckless” — A Republican-led Senate subcommitt

industrytechmeme

Hardware

Physical AI Takes the Wheel: How the World’s Robotaxi Leaders Are Building With NVIDIA Technologies

The global robotaxi market — physical AI’s first commercial breakthrough — is projected to reach $400 billion by 2035, with over 6 million commercial vehicles in operation as driverless fleets are alr

hardwarenvidia-blog

The Emergence of Social Science of Large Language Models

arXiv:2509.24877v3 Announce Type: replace Abstract: The social science of large language models (LLMs) examines how these systems evoke mind attributions, interact with one another, and transform huma

What Does MMLU Actually Measure? A Psychometric Audit of Difficulty Structure in Aggregate Benchmark Scores

arXiv:2609.09372v1 Announce Type: cross Abstract: Although MMLU is widely adopted as a benchmark for calibrating general AI capabilities, we psychometrically demonstrate that its aggregate score prima

When and Why LLM Causal Priors Help: Closed-Loop Prior Selection for Amortized Causal Inference

arXiv:2609.06941v1 Announce Type: new Abstract: Causal effect estimation asks how an outcome would change under an intervention, and medicine, economics, and public policy all treat it as a foundation

X2Streaming-ASR: wait when uncertain, emit when ready for streaming ASR

arXiv:2609.08672v1 Announce Type: cross Abstract: Streaming automatic speech recognition (ASR) for real-time voice agents and full-duplex dialogue must provide accurate partial transcripts with low co

Ambient @ EgoLongQA 2026: Distilling Long-Video perception into a Sub-2B Model

arXiv:2609.07154v1 Announce Type: new Abstract: We describe our entry to the EgoLongQA track of the Wearable-AI Challenge in ECCV 2026, which placed first in the <=2B parameter division with 0.8279 on

Closed-Loop Evaluation of Bird's-Eye-View Maps from Cross-View Transformers as Inputs to Behavior-Cloning Policies

arXiv:2609.05783v1 Announce Type: cross Abstract: In autonomous driving, Bird's-Eye View (BEV) representations provide a structured, top-down abstraction of the vehicle's surroundings and have become

safetyarxiv-cs-cv

Dual-Latent Memory Routing for Vision-Language Reasoning

arXiv:2609.05539v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) have recently made strong progress in vision-language reasoning, yet their performance often degrades as genera

Emergent Charging Coordination in Electric Delivery Fleets

arXiv:2609.07689v1 Announce Type: new Abstract: In electric delivery fleets, mid-shift charging is non-trivial: each vehicle must decide when, where and how much to charge to finish on time with batte

safetyarxiv-cs-lg

LatentMD: Benchmarking Markdown Boundary Failures in LLM-Generated Text

model-releasesarxiv-cs-lg

arXiv:2609.06993v1 Announce Type: cross Abstract: Large language models (LLMs) increasingly generate Markdown that is consumed by renderers, agents, code extractors, and structured downstream pipeline

Learning Acrobatic Flight from Preferences

model-releasesarxiv-cs-lg

arXiv:2508.18817v3 Announce Type: replace-cross Abstract: Preference-based reinforcement learning (PbRL) enables agents to learn control policies without requiring manually designed reward functions,

Local Ai

Map the Possibilities: Spatial Belief Fields for Language-Goal Aerial Navigation

arXiv:2609.05841v1 Announce Type: new Abstract: Language-goal aerial navigation requires an agent to local- ize a potentially unobserved target from relational instruc- tions and partial observations,

local-aiarxiv-cs-cv

NeuCME: Toward Dynamic Multimodal Continual Learning via Neural Combinatorics of Multiple Experts

arXiv:2609.07009v1 Announce Type: cross Abstract: Multimodal continual learning has recently shown great potential for developing agents with human-like intelligence by continuously learning new tasks

Physico-Geospatial Grounded Scene Interpretation for Mobile Robotics

arXiv:2609.06629v1 Announce Type: new Abstract: Recent advancements in deep learning allow robotic agents to interact with dynamic and unstructured environments. Of special interest is the integration

researcharxiv-cs-cv

Local Ai

PhysReal: Learning Real-World Deformable Object Physics via Hybrid Constitutive Modeling

arXiv:2609.07532v1 Announce Type: new Abstract: Learning physically plausible dynamics from visual observations is essential for interactive world models and embodied agents. However, modeling real-wo

local-aiarxiv-cs-ro

Hardware

PlayTrain: An Efficient Reinforcement Learning Framework for LLM-Generated Adaptable JavaScript Games

arXiv:2609.09059v1 Announce Type: new Abstract: While many video-game environments (VGEs) have played crucial roles in advancing reinforcement learning (RL), developing novel VGEs or modifying existin

hardwarearxiv-cs-lg

@sama Maybe. Or maybe this isn’t even an isolated incident, just like the Hugging Face incident wasn’t quite a one-off:

@sama Maybe. Or maybe this isn’t even an isolated incident, just like the Hugging Face incident wasn’t quite a one-off: BREAKING: OpenAI might have stolen another major proof. In a detailed Mastodon p

safetygary-marcus--x

What’s new in Microsoft Foundry: July and August 2026

model-releasesmicrosoft-foundry

Catch up on July and August in Microsoft Foundry: stable Hosted Agent and Toolbox SDK paths, Claude tools on Azure-hosted deployments, Model Router updates, local inference improvements, and the SDK m

enterprise-aicodexsoftware-delivery

Applications

1Password deploys Codex across engineering workflows

Human

1Password uses Codex across software delivery; its OpenAI case study reports 20.9% higher engineering productivity in a core cohort.

8 Sept 2026

Industry

Bed time reading 🥱

Bed time reading 🥱 We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, usin

industryemad-mostaque--x

8 Sept 2026

We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proo…

We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-ge

model-releasesopenai--x

8 Sept 2026

Compact Bellman-Grounded Cognitive Maps for Cost-Aware Navigation

arXiv:2609.05104v1 Announce Type: new Abstract: Biological agents navigate familiar environments not by re-solving routes for each new goal, but by reusing a learned map built once and read off as goa

researcharxiv-cs-ai

Design Docs Are All You Need: An AI-native Machine-Learning Performance Tool

arXiv:2609.05364v1 Announce Type: cross Abstract: Machine-learning performance modeling is a uniquely hostile terrain for long-lived software: the assumptions baked into today's abstractions are inval

// Design Docs Are All You Need // Banger paper from Google DeepMind, MIT, and colleagues. What a genuinely strange and interesting paper th…

// Design Docs Are All You Need // Banger paper from Google DeepMind, MIT, and colleagues. What a genuinely strange and interesting paper this one is. Here is the setup: They maintain a performance-mo

model-releasesdair-ai--x

Draft acceptance tweaks?

What can I do to get more from MTP if anything? It am I sitting at the expected range. Thanks /opt/llama.cpp/build/bin/llama-server --host 10.210.44.126 --port 11434 --model /opt/models/Qwen3.6-35B-A3