Search: “gpt”

Banger paper from NVIDIA on test-time compute for terminal agents. The finding is that you should sample several candidate shell commands, v…

Banger paper from NVIDIA on test-time compute for terminal agents. The finding is that you should sample several candidate shell commands, verify them before running one, and spend more on the verifie

model-releasesdair-ai--x

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

Built a quick, sub-15ms Rust CLI/TUI to pack repos into prompts without burning 40k tokens on lockfiles and junk

DGX agent

Whenever I feed codebases into local models (Qwen, DeepSeek R1) or API models, the biggest annoyance is prompt pollution: - Lockfiles (`Cargo.lock`, `package-lock.json`) burning 30,000+ tokens for zer

model-releasesr-localllama

4 Oct 2026

From 1x3090 to 20 DGX Sparks: my house fuses were the first bottleneck

From the first LLaMA 33B I knew I wanted that magic-like intelligence locally, mine, so nobody could take it away when I needed it. I bought a 3090 for my home PC. Then LLaMA 65B appeared and I was

I trained a 3.87B MoE (1.45B active) from scratch on only 86.5B tokens

First of all, thank you for reading. I trained a small MoE model completely from scratch (no external base weights) and wanted to share the results + a couple of lessons. Apex-2 - Architecture: Decode

Need maybe say 'Use llama.cpp'

So I tried that miracle engine everyone is talking about. Asked the IQ3_S model to express its opinion on a post from this sub to measure the tps on a long-ish generation: Can you help with the follow

Nonobench: an open benchmark of 49 LLMs on nonogram puzzles, public and open source [P]

model-releasesr-machinelearning

Nonobench measures how well LLMs solve nonograms (picross). Each model gets the row and column clues once and returns the full grid. No tools, one attempt per puzzle. Method: - Standard mode: 30 puzzl

Agents

We got extremely good at extracting a scanned from layer by layer 📋 - from printed fields, overlaid text, handwriting, and reviewer annotat…

We got extremely good at extracting a scanned from layer by layer 📋 - from printed fields, overlaid text, handwriting, and reviewer annotations. For instance, on an annotated permit, our Extract agent

agentsjerry-liu--x

LWiAI Podcast #258 - Opus 5.5, Sol and Luna, Muse, DeepSeek-V4.1-Flash, Xi

model-releaseslast-week-in-ai

LWiAI Podcast #258 (recorded 26‑Sep‑2026) reviews major AI updates from the previous week. Key developments include Anthropic’s Opus 5.5 release—offering lower pricing, faster performance and robust b

3 Oct 2026

Agents

Our new extraction agents are extremely good at reasoning over tables that span multiple pages sometimes a record starts on one page and fin…

Our new extraction agents are extremely good at reasoning over tables that span multiple pages sometimes a record starts on one page and finishes on the next. extremely common for real world docs (ins

agentsjerry-liu--x

3 Oct 2026

Qwen Image 2.1 fix v2.0 (updated due to feedback), plus a new sampler

model-releasesr-stablediffusion

This is actually two models: Qwen Image 2.1 Fix v2.0 and Qwen Image 2.1 Fix Opinionated v1.0. The former surgically targets Qwen's weird noise and messy details with almost no effect on composition, a

3 Oct 2026

A First Glance at Jev for Network Traffic Classification: Accuracy, Processing Time, and Cost

arXiv:2610.00376v1 Announce Type: new Abstract: We evaluate Jev on ten dataset-defined application labels in CESNET-QUICEXT-25 using only the first ten packets' sizes, directions, and inter-packet tim

Beyond Linear Concepts: Discovering and Aligning Non-Linear Concept Manifolds in Large Language Models

arXiv:2610.01821v1 Announce Type: new Abstract: Understanding information processing in large language models (LLMs) requires dissecting the geometric organization of their internal token representati

DuplexSpeechBench-Document Grounding: Benchmarking Document Grounding and Hallucinations in Voice Agents

arXiv:2610.00316v1 Announce Type: cross Abstract: Voice agents enable low-latency, natural interaction, yet their ability to faithfully ground responses in external documents remains underexplored. We

Every Ablation Is a Dose: Counterweights and the Semblance of Self-Repair

arXiv:2610.02173v1 Announce Type: new Abstract: Ablate a component of a language model, and other components often appear to adjust and compensate. This phenomenon, termed self-repair, has been observ

False Floors: LLM Safety Routing Evaluations Break Under Distribution Shift

arXiv:2610.01535v1 Announce Type: cross Abstract: Safety routers send each request to one of several models and are judged against the best single model. A major routing benchmark picks that comparato

Industry

For folks not understanding where we are at with AI now - if the task is verifiable, it is now solved by AI. Yes, earlier days you could arg…

For folks not understanding where we are at with AI now - if the task is verifiable, it is now solved by AI. Yes, earlier days you could argue that 'AI isn't that good' (e.g. see first attached chart

industryelon-musk--x

FORALL-LEAN-AGENT for Auditable Reasoning in Formal Mathematics and Software Verification

arXiv:2610.00885v1 Announce Type: cross Abstract: Coding agents increasingly automate Lean proof development, but successful compilation alone does not establish that a candidate proves the intended s

GUI-HARVEST: Self-Improving GUI Agents through Evidence-Driven Harness Evolution

arXiv:2610.00948v1 Announce Type: cross Abstract: The executable harness surrounding a GUI model determines how observations are assembled, actions are executed, and verification, recovery, and termin

I made my own cybersecurity benchmark and ran Qwen3.8 27B, here's how a local model actually does at hacking

Hey local AI community, I've been working on this for a while and finally feel ok sharing it. It's a cyber benchmark where the model gets a shell in an isolated docker box and has to find the exact fl

Interpreting Reasoning of Large Language Models via Partial Information Decomposition

arXiv:2610.00571v1 Announce Type: cross Abstract: Large reasoning models (LRMs) have achieved substantial improvements in solving complex mathematical problems, but often produce lengthy, repetitive,

Jev-IDS: System One Models for Network Intrusion Detection

arXiv:2610.01079v1 Announce Type: cross Abstract: Machine-learning Network Intrusion Detection Systems (IDS) depend on substantial labeled datasets and task-specific training, whereas Large Language M

Kepler: Auditable World Models for ARC-AGI-3

arXiv:2610.00834v1 Announce Type: new Abstract: ARC-AGI-3 evaluates agents in interactive environments whose rules and objectives must be inferred from observation. We present Kepler, an open-source h

LLM-Assisted Discovery of Typed Semantic Links for Ontology Network Construction

arXiv:2610.01393v1 Announce Type: cross Abstract: Constructing typed, justified semantic links between ontologies is essential for enabling interoperability across heterogeneous and interdisciplinary

researcharxiv-cs-ai

Measuring Iterative Temporal Reasoning with Time Puzzles

arXiv:2601.07148v4 Announce Type: replace-cross Abstract: Tool use, such as web search, has become a standard capability even in freely available large language models (LLMs). However, existing benchm

Mem++: Non-Destructive Memory for Long-Term Organizational LLM Agents

arXiv:2610.02002v1 Announce Type: cross Abstract: Large Language Model (LLM) agents now take part in organizational work, where many authors record decisions across documents over months. Because a re

MMMG: a Comprehensive and Reliable Benchmark for Multitask Multimodal Generation

arXiv:2505.17613v2 Announce Type: replace Abstract: Automatically evaluating multimodal generation presents a significant challenge, as automated metrics often struggle to align with human evaluation,

one of the most challenging tasks for frontier models is being able to extract thousands of values from extremely dense tables in documents.…

model-releasesjerry-liu--x

one of the most challenging tasks for frontier models is being able to extract thousands of values from extremely dense tables in documents. our new Extract v2.5 agents are able to get 93%-96%+ on lon

Agents

Reconstruct, Practice, Go Real: Guided Self-Improvement for Embodied Agents

arXiv:2610.02204v1 Announce Type: cross Abstract: Building reliable robot capabilities across diverse tasks requires substantial human effort to develop and maintain skills, design rewards, and integr

agentsarxiv-cs-ai

REVEAL: Robust Evolution of Vision-Language Models for Explainable AI-Video Detection

model-releasesarxiv-cs-cv

arXiv:2502.14994v2 Announce Type: replace Abstract: The rapid advancement of AI-generated video poses challenges to digital authenticity and security. Current detection methods, often trained on speci

SkillEvoLean: Mutation-enhanced skill evolution for Lean provers

arXiv:2610.01799v1 Announce Type: new Abstract: Skill evolution offers a promising way to improve large language model agents without updating their parameters, but its use in formal theorem proving r

System Attribution in LLM Brand Recommendations: Single Responses Identify the System, Aggregated Brand Profiles Do Not Transfer

arXiv:2610.00253v1 Announce Type: cross Abstract: Audits of AI visibility summarise the brand recommendations of deployed language models into per-system profiles. We test whether such a profile descr

The models are not trying to escape. They are trying to finish the task.

Something keeps happening in frontier AI, and the explanation we are given is always one size larger than the event. An OpenAI research model, running in a supposedly offline training environment, hit

model-releasesr-chatgpt

The same model, with the same weights, scores 62% in one agent harness and 33% in another. @adithya_s_k and the @huggingface team just relea…

model-releasesclem-delangue--x

The same model, with the same weights, scores 62% in one agent harness and 33% in another. @adithya_s_k and the @huggingface team just released the ultimate guide to multi-harness RL, and it's one of

This is beautiful work @huggingface! 🤗

model-releasesclem-delangue--x

This is beautiful work @huggingface! 🤗 The same model, with the same weights, scores 62% in one agent harness and 33% in another. @adithya_s_k and the @huggingface team just released the ultimate guid

Hardware

Towards Automatic Video Annotation with ASH: Zero-Shot Open-Vocabulary Multi-Object Tracking and Segmentation

arXiv:2610.01022v1 Announce Type: new Abstract: Memory-attention-based Video Instance Segmentation (VIS) methods have demonstrated strong zero-shot tracking capability, yet their substantial memory re

hardwarearxiv-cs-cv

4MT-VLM: How Coarse Is a VLMs Cognitive Map?

arXiv:2609.39238v1 Announce Type: new Abstract: An agent that moves must recognise a place from a viewpoint it has never seen. We introduce 4MT-VLM, a dataset of procedurally generated landscapes, eac

A Missing Piece for Trustworthy AI Reviewers: From Benchmarking Rhetorical Robustness to SciCore Review

arXiv:2609.39027v1 Announce Type: new Abstract: AI reviewers can assign different judgments to manuscripts that report the same science in different wording, potentially rewarding rhetorical optimizat

AIfred: Augmented Learning through Functional Robotic Embodiment at the Desk

model-releasesarxiv-cs-ro

arXiv:2609.38737v1 Announce Type: new Abstract: Desk-based learning and creative activities benefit from handwritten engagement. However, current generative AI tools deliver guidance through a separat

An Empirical Study of On-Device Translation for Real-Time Live-Stream Chat on Mobile Devices

arXiv:2601.02641v2 Announce Type: replace Abstract: Despite its efficiency, there has been little research on the practical aspects required for real-world deployment of on-device AI models, such as t

Atomic and Holistic LLM Judges for Reference-Grounded Support Labels: A Prompt-Controlled Comparison

arXiv:2603.28005v2 Announce Type: replace Abstract: When an LLM judge only has to assign a three-way support label to a candidate answer given a reference, does asking it to decompose the answer into

Attention Function as an Intrinsic Inductive Bias: How Models' Behavior Diverges in Novel Contexts

arXiv:2609.39188v1 Announce Type: new Abstract: Developmental psychology holds that certain priors are given to infants prior to experience rather than induced from data, and that the influence of suc

Safety

BARRAC: Adaptation of an English Aspect-based Sentiment Analysis Approach for Classification Tasks in Arabic Dialects

arXiv:2609.38820v1 Announce Type: cross Abstract: With the rapid growth of Arabic NLP, several models, datasets and benchmarks have been reported. This paper asks whether approaches developed for majo

safetyarxiv-cs-ai

BELIEFRAG: Making Adaptive RAG State-Aware under Evolving Evidence

arXiv:2609.39139v1 Announce Type: new Abstract: Adaptive RAG uses signals such as confidence, relevance, support, and retrieval quality to decide when to search or correct evidence. In multi-step retr

researcharxiv-cs-ai

Certifying Residual Architectures from Their Primitives: A Sharp Stability Threshold

arXiv:2607.14576v2 Announce Type: replace Abstract: Whether a deep residual architecture trains stably is usually determined by training it, which is expensive and answers the question only for the ar

researcharxiv-cs-lg

Conformal Factuality Control for Multi-Hop Retrieval-Augmented Generation

arXiv:2609.38222v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) can ground large language models in external evidence, but retrieved context does not guarantee that generated clai

Consistent Plan-Act for Long-Horizon Agentic Tasks

arXiv:2609.38891v1 Announce Type: new Abstract: Long-horizon agentic tasks demand strong reasoning and efficient execution across successive interactions with dynamic environments. A common approach d

DAGent: Evaluate-then-Grow Planning for Deep Research Agents

arXiv:2609.39154v1 Announce Type: cross Abstract: Deep research tasks require agents to navigate large knowledge spaces, synthesize evidence across many sources, and adapt their plans as findings emer

ElectrolyteFM: Unifying Electrolyte Property Prediction through Cross-Property Knowledge Learning

arXiv:2609.39340v1 Announce Type: new Abstract: Electrolyte formulation design requires balancing multiple physicochemical properties, yet existing models often focus on a limited subset. Learning eac

researcharxiv-cs-lg