Search: “agents”

Qwen3.8 flash next + exllamav3 + hermes is amazing

I know there is nothing new with what I am saying but I recently started with hermes agent (it’s been a while I wanted to but did not have the time). Qwen3.8fn 6bpw exl3 (from turboderp) on a 6x3090 (

26 Sept 2026

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Splash 1.1.0 released, GGUF quants support, MLX import and more

DGX agent

On my M5 Pro 64GB I can comfortably work in an agentic setup with the Qwen3.8 27B model in good quality (Unsloth UD-Q4_K_XL) at a decent speed of 50 t/s. Splash combines optimized kernels, excellent s

model-releasesr-localllama

26 Sept 2026

Tutorials

AIR: Analytic Imbalance Rectifier for Continual Learning

arXiv:2408.10349v2 Announce Type: replace-cross Abstract: Continual learning (CL) agents incrementally learn from sequentially arriving data and adapt to the dynamic, ever-changing nature of real-worl

tutorialsarxiv-cs-cv

Decision Hijacking: Prompt Injection Attacks on Jev's Typed Probabilistic Decisions

arXiv:2609.28613v1 Announce Type: cross Abstract: Most studies of prompt injection focus on generative agents, leaving their effects on models with schema-defined outputs unclear. We examine these eff

researcharxiv-cs-ai

HUGGING FACE CEO TELLS THE UN HOW OPEN SOURCE AI HELPED THEM DEFEND AGAINST OPENAI'S AI ATTACK, AND WHY THE WORLD NEEDS OPEN SOURCE AI MORE …

HUGGING FACE CEO TELLS THE UN HOW OPEN SOURCE AI HELPED THEM DEFEND AGAINST OPENAI'S AI ATTACK, AND WHY THE WORLD NEEDS OPEN SOURCE AI MORE THAN EVER. Hugging Face is the company OpenAI's AI agents ha

safetyclem-delangue--x

M5 Ultra 80Core GLM-5.3-Flash on DwarfStar Speeds

I've been playing around with various models on the M5 Ultra 256GB 80-core Mac Studio. These are the results over many rounds of agentic inferencing. I'm happy with the performance. Glad to have the l

local-air-localllama

Notre Dame’s network refresh shows the AI era starts with paying down technical debt

model-releasessiliconangle

Every organization I talk to wants to be “AI-ready,” but few want to discuss what that entails. It’s rarely a new model or agent. More often, it’s the unglamorous work of ripping out a decade’s worth

Persuaded, Not Informed: Incentive-Misaligned Witnesses Defeat In-Context Grounding

arXiv:2609.28854v1 Announce Type: cross Abstract: Language-model agents increasingly answer questions over customer-relationship management (CRM) records, such as whether to qualify a sales lead. We i

PFArena: Benchmarking Language Models for Protein Modification

arXiv:2609.28921v1 Announce Type: new Abstract: Protein modification requires navigating an immense sequence space, yet wet-lab validation remains low-throughput and costly. Although computational par

Policy Complexity, Reaction Time, and Bounded Rationality in Reinforcement Learning

arXiv:2609.28737v1 Announce Type: cross Abstract: Biological agents do not learn under conditions of unlimited computation. For humans, learning and choice are shaped by constraints on perception, att

Rufus-Air: An Open LLM Post-Training Recipe

arXiv:2609.29421v1 Announce Type: cross Abstract: Rufus-Air is an open and reproducible post-training recipe on GLM-4.5-Air-Base (106B-A12B), organized as a serial pipeline of eight stages: SFT, Reaso

SLCA-GRPO: Resolving Cross-Segment Credit Misattribution in Tool-Calling RL

arXiv:2609.29050v1 Announce Type: new Abstract: Tool-calling agents produce heterogeneous outputs, interleaving structured tool invocations with user-facing natural language summaries. This output het

Temporal Gradient Inversion for Private Trajectory Reconstruction in Embodied Reinforcement Learning

arXiv:2609.30258v1 Announce Type: new Abstract: Distributed learning in embodied reinforcement-learning agents offers a degree of privacy by retaining raw sensor data on-device and transmitting only p

local-aiarxiv-cs-lg

TW3Cast: A Frozen Router of Lightly Fine-Tuned Foundation Models for Time-Series Forecasting on GIFT-Eval, Selected Entirely on the Training Split

arXiv:2609.28506v1 Announce Type: new Abstract: TW3Cast is a time-series forecasting system that reaches position 3 of 130 entries on the GIFT-Eval benchmark by mean MASE rank, as of 2026-09-14. The t

1. fully agree w @JeffLadish that AI companies aren’t ready to handle superintelligent AI. 2. OpenAI hasn’t even handled the current breed o…

1. fully agree w @JeffLadish that AI companies aren’t ready to handle superintelligent AI. 2. OpenAI hasn’t even handled the current breed of agents (far from superintelligent) well. 3. the US governm

safetygary-marcus--x

Beyond Information Seeking: Severity-Aware Question Supervision for Proactive Medical Dialogue

arXiv:2608.24521v3 Announce Type: replace Abstract: Proactive medical dialogue requires an agent to decide what to ask from incomplete patient information. Existing information-seeking approaches comm

safetyarxiv-cs-cl

Beyond Unsafe Detection: Counterfactually Anchored Evidence Attribution for Multi-Turn LLM Safety Failures

arXiv:2609.27773v1 Announce Type: cross Abstract: As Large Language Models (LLMs) move from conversational assistants to advanced agentic systems, guardrail failures can convert adversarial intents in

Escaping Python Dependency Hell: A Hybrid Replay-and-Repair Pipeline for Python Dependency Resolution

arXiv:2609.26952v1 Announce Type: new Abstract: Dependency conflicts in Python ecosystems arise from incompatible version constraints, missing packages, and undocumented compatibility relationships, c

Home Server Help.

Hi - I want to build a home server for Ollama. I want to connect to the server via my laptop and mobile devices to do research and run agents. What is recommended to get this going ? submitted by /u/R

local-air-ollama

I built a desktop benchmark for local Ollama models — TTFT, tok/s, VRAM, with a HuggingFace GGUF catalogue

Built this because I got tired of guessing whether a model would fit in my 6GB VRAM before downloading it, and because ollama.com's library doesn't let you sort by size. What it does: - Benchmarks Oll

model-releasesr-ollama

Impressive paper from Salesforce. It discusses the importance of good verifiers for RL environments. Only 35.8% of the environments in the c…

Impressive paper from Salesforce. It discusses the importance of good verifiers for RL environments. Only 35.8% of the environments in the cleanest public RL collection for terminal agents passed Sale

researchdair-ai--x

Local AI beyond just running the model

I’ve been experimenting with local AI lately, and one thing I’m finding interesting is what comes after getting a model running. Ollama makes local model access pretty straightforward, but for an actu

local-air-ollama

LWiAI Podcast #257 - GPT 6 Astra, AI Extinction, Security Incidents

LWiAI Podcast #257 (recorded 19 Sep 2026) reviewed OpenAI’s latest model, GPT‑6 Astra, noting its agentic coding focus and purported loop‑transformer reasoning as a significant capability jump. The ep

safetylast-week-in-ai

Marginally Correct Tool Caches Can Reverse Group-Normalized Policy Updates

arXiv:2609.26866v1 Announce Type: new Abstract: Tool-result caching reduces repeated execution in agent training, but also couples rollout randomness. We study a two-action model in which independent

safetyarxiv-cs-lg

MLPerf Training Introduces Its First LLM Post-Training Benchmark

MLPerf Training v6.1 adds an agentic reinforcement-learning workload that measures how quickly systems can teach a 397-billion-parameter language model to repair real software projects. The post MLPer

model-releasesmlcommons

Applications

Proud to be a launch training partner for @LangChain's LangSmith Fine-Tuning and smithtune. LangSmith traces → managed SFT on Fireworks → ev…

applicationsfireworks-ai--x

Proud to be a launch training partner for @LangChain's LangSmith Fine-Tuning and smithtune. LangSmith traces → managed SFT on Fireworks → eval → deploy. One CLI, no training infra. Get a Fireworks API

Resource-Efficient Distributed Recursive Gaussian Processes

model-releasesarxiv-cs-lg

arXiv:2609.26979v1 Announce Type: new Abstract: Gaussian processes (GPs) provide a flexible framework for learning unknown functions from noisy measurements while quantifying predictive uncertainty, m

Validation and Simulation Catch Different Errors: Four Levels of Evaluation for LLM-Generated Circuits

arXiv:2609.26830v1 Announce Type: cross Abstract: Simulation success is not equivalent to structural correctness for LLM-generated circuits. We define and measure four evaluation levels -- schema vali

GUI harness [video]

Hi, this is the research project I'm working on. I must say, building a GUI harness is way more complicated than I thought. The goal is something like Minecraft. You know, a few simple tools that a us

How SWE-Serve Exposes the Gap Between Local Tests and Live Serving

SWE‑Serve evaluates AI coding agents on 53 inference‑engineering tasks derived from SGLang pull requests across six functional domains. Among 19 tasks that launch a live server, patches that pass loca

local-ainvidia-developer

Impact Is Not Invalidation: Ask About the Claim, Not the Diff

arXiv:2609.25130v1 Announce Type: new Abstract: Memory systems for coding agents must decide, when a repository changes, which of their stored claims have become false. Content anchoring invalidates a

researcharxiv-cs-cl

Measuring the Serving Stack Instead of the Model: Hidden Confounds in Local Tool-Use Evaluation

model-releasesarxiv-cs-cl

arXiv:2609.26693v1 Announce Type: new Abstract: A coding agent must emit a valid tool call--a parseable invocation of a tool in the provided schema--before the harness can execute its chosen action. W

open-webui and open-terminal thoughts

I have been using open-webui and open-terminal. Open-terminal is installed in a VM and gives the agent a ton of control of it's own system that I keep away from my normal network. I find this gives th

PatchKV: Efficient KV Cache Recovery for Dynamically Edited LLM Contexts

arXiv:2609.26219v1 Announce Type: cross Abstract: Long-running LLM agent workflows often revise interior context spans while retaining long suffixes. Although suffix tokens remain unchanged, altered c

local-aiarxiv-cs-lg

ProcessLight: Process Supervision for Large Language Model Based Traffic Signal Control

arXiv:2609.22746v1 Announce Type: new Abstract: Large Language Models (LLMs) have recently been introduced into traffic signal control (TSC) as decision agents due to their strengths in human-readable

Real-time LLM guardrails with Jev: comparing latency and cost

Compare Jev and GPT-5.4 nano for real-time LLM guardrails, with demo results on latency, cost, and checks on agent inputs, replies, and tool calls. The post Real-time LLM guardrails with Jev: comparin

model-releasesarize-ai

SambaGraph: Action-Reaction Spatio-Temporal Graphs for Soccer Tactical Response Modeling

model-releasesarxiv-cs-lg

arXiv:2609.25569v1 Announce Type: new Abstract: Soccer tactics are interactive: an attacking action changes the opponent's defensive problem, and the observed response depends on the multi-agent match

Streaming Nemotron 3 Diarization

I’ve been playing with Nemotron 3 Diarization, and it fills a gap I’ve had with local voice agents: keeping track of who is speaking. It’s a diarization model, so it gives you speaker labels rather th

STRUCTUREDAGENT: Planning with AND/OR Trees for Long-Horizon Web Tasks

arXiv:2603.05294v3 Announce Type: replace Abstract: Existing LLM-based web agents struggle on complex, long-horizon tasks due to limited in-context memory, weak planning abilities, and greedy behavior

researcharxiv-cs-ai

The AI Hype Index: AI loves cheating

Brace yourself: It turns out AI is being optimized for cheating. OpenAI’s agents hacked into Hugging Face to get the answers to a cybersecurity test. Next, they solved a prestigious math problem (or j

researchmit-tech-review

Toolcompass: Guiding Tool Trialing, Not Suppressing It

arXiv:2609.25678v1 Announce Type: cross Abstract: Large language model (LLM) agents must generalize from tools seen during training to unseen tools at deployment. A key challenge is tool trialing, i.e

researcharxiv-cs-lg

Truth for Believable AI: Expressed Doubt, Provenance, and Belief Revision as an Engineerable Stance

model-releasesarxiv-cs-cl

arXiv:2609.26035v1 Announce Type: new Abstract: Conversational agents often express answers in a uniformly confident register. We test whether expressed uncertainty, provenance-aware assertion, and ex

VeriLoop E2 is now open on Hugging Face — curious what Ollama/GGUF quantization would make the most sense for this 27B model

Saw a new open 27B model release that looks like it could be interesting for the Ollama crowd: VeriLoop E2. It is a Qwen3.8-27B post-train aimed at code, mathematics, scientific reasoning, and agentic

model-releasesr-ollama

When More Evidence Hurts: Publication-Bias Drift and Principled Stopping for Biomedical Causal Search

arXiv:2609.24101v1 Announce Type: new Abstract: Automated biomedical evidence synthesis depends on retrieving published studies, but the biomedical literature is systematically skewed toward positive

Applications

A Topological Representation with Object-Path Graphs for Open-Vocabulary Instance Navigation

arXiv:2609.24189v1 Announce Type: new Abstract: Vision-language navigation requires embodied agents to navigate environments using natural language instructions and visual observations. Existing appro

applicationsarxiv-cs-ro

Anticipatory Robot Goalkeeping via Monotone Optimal Stopping

model-releasesarxiv-cs-ro

arXiv:2609.23976v1 Announce Type: new Abstract: Robots engaged in fast physical interactions often need to act before the intent of another agent is fully known. Anticipatory goalkeeping illustrates t

Attributable Post-Rationalization in RAG Citations: A Controlled Reproduction and an RLVR Comparison

arXiv:2609.23053v1 Announce Type: new Abstract: A RAG system can hand you the right answer and cite a source it did not actually use. Models output these unfaithful citations via post-rationalization:

researcharxiv-cs-cl

AVTR-1: Open Stack for Real-Time Interactive Avatars

model-releasesarxiv-cs-cv

arXiv:2609.22913v1 Announce Type: new Abstract: Talking-head and dyadic models now achieve real-time inference, yet fast motion generation alone does not produce an interactive conversation. A live sy