Search: “models”

WebGrader: Training LLMs for Web Development with Self-Evolving Programmatic Grader

arXiv:2608.06474v1 Announce Type: new Abstract: Large language models increasingly generate complete websites from natural-language descriptions, and reinforcement learning has become a central approa

10 Aug 2026

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

AMD llama.cpp: reducing MTP buffer overhead gave me 64K → 149K context for Qwen 27B

DGX agent

Available context length with and without the patch: Model: QWEN 27B ROCm stock patched Vulkan stock patched IQ4_XS Pure, single 16GB GPU 19.456 76.032 68,352 78,592 Q6_K_L on 16GB + 12GB 64,256 149,2

model-releasesr-localllama

9 Aug 2026

DeepSeek v4 Flash 0731 locally on CPU

After seeing the benchmark results for the full release of DS v4 Flash 0731, I replaced my 2 x 16GB DDR4 ram sticks with 2 x 32GB DDR4 ram sticks to get a max supported of 128 GB RAM, in hope to be ab

M3 16GB running Ollama (Qwen 9B) is extremely slow (10-12 mins per task). Am I doing something wrong?

Hey everyone, I constantly see high praise for M3 and M4 Macs for local LLM inference, even the base/16GB models. However, my experience has been quite different, and I'm trying to figure out if I hav

model-releasesr-ollama

Quoting Claude Opus 5 system prompt

model-releasessimon-willison

Claude Fable 5 and Claude Mythos 5 were first released on June 9, 2026. On June 12, 2026, Anthropic suspended access to both models to comply with U.S. Department of Commerce export controls; the Depa

Underestimated budget solution: radeon 780m iGPU

There are so many posts where people complaining about high prices and asking for solution <= 1000 EUR. So, there is one solution to consider: PC/mini PC/laptop on Ryzen 7 260/Ryzen 9 8945HX/etc CPU w

Claude Code in 9 lines python

I was wondering what a minimal coding agent implementation would look like that can be used like Claude Code or Codex Not feature-by-feature of course but basically stripping everything out that is no

8 Aug 2026

Currently serving 200 tps+ output speed for DeepSeek-V4-Flash on Ollama's cloud with zero data retention (ZDR). Have an amazing weekend 🫡

Currently serving 200 tps+ output speed for DeepSeek-V4-Flash on Ollama's cloud with zero data retention (ZDR). Have an amazing weekend 🫡 DeepSeek-V4-Flash-0731 is now fully rolled out as the new defa

model-releasesollama--x

8 Aug 2026

LiteParse can now extract structured data from your PDF in milliseconds: ✅ checkbox states ✅ annotations ✅ vector graphics ✅ word-level boun…

model-releasesjerry-liu--x

LiteParse can now extract structured data from your PDF in milliseconds: ✅ checkbox states ✅ annotations ✅ vector graphics ✅ word-level bounding boxes It is the most comprehensive, accurate (and fast)

8 Aug 2026

Local Ai

Activity Frames: Deterministic Screen-Activity Compilation for Agent Memory and Replay

arXiv:2608.05784v1 Announce Type: new Abstract: Computer-use agents pay full frontier inference to re-derive routines their user has already performed, because an agent's memory today records what the

local-aiarxiv-cs-ai

Align-RAG: Alignment Is All You Need for TSFM In-Context Learning

arXiv:2608.05571v1 Announce Type: new Abstract: Retrieval-augmented forecasting promises to adapt frozen Time Series Foundation Models (TSFMs) to new domains without fine-tuning, but recent methods ty

Autonomous Research Agents: A Survey of AI Scientists and the Verification Gap

arXiv:2608.05179v1 Announce Type: cross Abstract: Large language model (LLM) agents are increasingly used across the scientific research lifecycle: ideation, literature search, experiment design and e

b10312

model-releasesllama-cpp-releases

server: (router) do not evict busy models (#26567) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFram

b10326

model-releasesllama-cpp-releases

tts: account for the vocoder pass in the timings line (#26733) get_output runs the waveform work the pipeline defers to it, from a single trailing window to a full pass depending on the model. Measuri

Beyond Rotations: AuroOFT for Expressive Quantized Orthogonal Fine-Tuning

arXiv:2608.05253v1 Announce Type: new Abstract: Quantized orthogonal fine-tuning (qoft) enables parameter-efficient adaptation of low-bit language models by learning structured activation rotations be

CASCADE: An Agentic Regulatory Network Framework for Patient-Data-Validated Downstream Perturbation Prediction

arXiv:2608.05359v1 Announce Type: new Abstract: CASCADE is an agentic framework that predicts downstream transcriptional effects of gene perturbation from precomputed ARACNe regulatory networks, expos

ChainClaw: A Layered Agent Framework for Reliable On-Chain Execution

arXiv:2608.05790v1 Announce Type: new Abstract: General-purpose large language model agents have achieved strong performance on tool-augmented tasks, yet they rely on assumptions break down in blockch

Local Ai

ConceptADapt: Concept-guided Adaptive Feature Reconstruction with Dynamic Attention for Few-Shot Industrial Anomaly Detection

arXiv:2608.05743v1 Announce Type: new Abstract: Few-shot industrial anomaly detection (FS-IAD) focuses on detecting and localizing visual defects in industrial inspection during the cold-start phase,

local-aiarxiv-cs-cv

ConWriter: Transition-Constrained Stateful Long-Form Story Generation with Lightweight Neuro-Symbolic Consistency Control

arXiv:2608.05169v1 Announce Type: new Abstract: Long-form story generation requires models to preserve narrative consistency across extended contexts, yet existing prompting-based methods often accumu

Local Ai

Curia-MAE: Multi-Modal Multi-Anatomy MAE Pre-Training for 3D Medical Image Segmentation

arXiv:2608.05844v1 Announce Type: new Abstract: Radiology foundation models learn transferable representations that can be adapted to new tasks by training only small layers on top of a frozen encoder

local-aiarxiv-cs-cv

Enhancing Anomaly Resilience in Research Networks: A Large-Scale Forecasting Benchmark for Dynamic Security Baselining

arXiv:2608.05605v1 Announce Type: cross Abstract: Research and Education Networks (RENs) serve as critical infrastructure for scientific discovery, yet they face a unique security paradox: their norma

Enhancing Social Intelligence in LLMs with Hierarchical Reasoning and Utterance-Level Goal Rewarding

arXiv:2608.05832v1 Announce Type: new Abstract: Large language models (LLMs) excel in structured tasks but struggle with dynamic social interactions, where success requires long-term goal coordination

Research

From Continuous Predictors to Clinical Thresholds: Early Evidence on Performance Trade-offs of Guideline-Based Categorisation for Ischaemic Stroke Outcome Prediction

arXiv:2608.05203v1 Announce Type: new Abstract: Machine learning models achieve strong predictive accuracy for 90-day outcome prediction in acute ischaemic stroke, yet clinical adoption is limited by

researcharxiv-cs-ai

In this episode of @wandb's Gradient Dissent , @l2k and Fireworks CEO @lqiao discuss why she believes the industry is at a turning point... …

model-releasesfireworks-ai--x

In this episode of @wandb's Gradient Dissent , @l2k and Fireworks CEO @lqiao discuss why she believes the industry is at a turning point... ... one that calls for more open intelligence, not less. The

Iterate or Widen? When Test-Time Refinement Helps LiDAR Scene Completion: A Controlled Study of Evidence Geometry, Training Coverage, and Compute

arXiv:2608.06014v1 Announce Type: new Abstract: Should a completion model spend extra test-time compute by iterating, or spend a similar parameter budget on a wider one-shot predictor? The answer is e

LLM Inference Under Bursty Workload Distribution: Modifying the WAIT Algorithm

arXiv:2608.06135v1 Announce Type: new Abstract: Large Language Models (LLMs) such as ChatGPT and Claude are widely used for information retrieval and problem-solving. Recent work has focused on improv

M^3R-Bench: A Unified Benchmark for Evidence-Grounded Multimodal Metaphor Understanding

arXiv:2608.05817v1 Announce Type: new Abstract: Metaphor enables the understanding of abstract concepts through cross-domain mappings while conveying affective attitudes. In multimodal scenarios, visu

MAC 2026: Advancing Micro-Action Analysis Towards Fine-Grained Understanding

arXiv:2607.16284v2 Announce Type: replace Abstract: Micro-Actions (MAs) are subtle and spontaneous human behaviors that provide important non-verbal cues in social interaction and affective communicat

Operating Multi-Node Full Fine-Tuning on NVIDIA B300: A Field Report on Telemetry-Based Triage, Negative Results, and Operational Hardening

arXiv:2608.05944v1 Announce Type: cross Abstract: We report operational experience full-fine-tuning a 32.76B-parameter dense model (Qwen3-32B) on 16 x NVIDIA B300 (two nodes, FSDP / ZeRO-3) -- among t

SearchAuditor: Auditing and Attributing Failures in Long-Horizon Search Agents

arXiv:2608.05212v1 Announce Type: new Abstract: Deep search agents tackle challenging questions through long-horizon web interactions, a process that is both complex and fragile: small reasoning error

Stochastic Parrots or Singing in Harmony? Testing Five Leading LLMs for their Ability to Replicate a Human Survey with Synthetic Data

arXiv:2603.00059v3 Announce Type: replace-cross Abstract: How well can AI-derived synthetic research data replicate the responses of human participants? An emerging literature has begun to engage with

TAU-Bench: From Anomaly Instance Tracking to Fine-Grained Video Anomaly Understanding

arXiv:2608.05699v1 Announce Type: new Abstract: Humans understand anomalous events through a coherent perceptual process in which they identify the focal instance, follow its behavior as the event unf

Safety

Visual Grounding in Zero-Shot Vision-Language Control

arXiv:2608.06154v1 Announce Type: cross Abstract: Vision-language models (VLMs) are increasingly used as zero-shot controllers, but successful trajectories do not necessarily show that decisions are g

safetyarxiv-cs-ai

What Drives Test-Time Adaptation for CLIP? A Controlled Empirical Study from an Update Perspective

arXiv:2606.14299v2 Announce Type: replace Abstract: Vision-Language Models (VLMs) such as CLIP have become a standard backbone for open-vocabulary recognition, yet their zero-shot predictions remain v

Who Checks the Citations? Benchmarking Legal Hallucination Detection

arXiv:2606.21155v2 Announce Type: replace Abstract: Attorneys, judges, and pro se filers increasingly use AI to draft legal documents, yet these tools frequently fabricate citations. Despite predictio

Research

A Mechanistic Analysis of Transformers for Dynamical Systems

arXiv:2512.21113v2 Announce Type: replace Abstract: Transformers are increasingly adopted for modeling and forecasting time-series, yet their internal mechanisms remain poorly understood from a dynami

researcharxiv-cs-lg

A Survey of Agent Memory in the Second Half: Towards Self-Evolving and Long-Horizon Agents

arXiv:2602.06052v4 Announce Type: replace-cross Abstract: Research in artificial intelligence is shifting from model innovations and benchmark scores towards problem definition and rigorous real-world

Active-SWE: Benchmarking Coding Agents for Proactive Bug Fixing without Issue Reports

arXiv:2608.04682v1 Announce Type: cross Abstract: Coding agents powered by large language models (LLMs) are increasingly adopted in software engineering (SWE) scenarios, capable of fixing a specific b

Adaptive Finite-Budget Training for CVaR Risk-Aware Q-Learning

arXiv:2608.04305v1 Announce Type: new Abstract: Risk-aware Q-learning (RaQL) provides a model-free, two-timescale estimator for dynamic risk objectives, but its finite-budget behavior remains fragile:

Agentic Future Ready With BigQuery: Continually Improving Price-Performance, Zero Effort

model-releasesgoogle-cloud-ai

In the modern data landscape, query performance tuning and managing system price-performance is challenging, especially as the number of agentic workloads increase. Even for experienced developers and

Auto-fit vs tuned MoE offload: 564 → 1330 pp tok/s, unchanged decode (Qwen3.6-35B-A3B Q6 / RTX 3090)

TL;DR: On a Qwen3.6-35B-A3B Q6 setup sized for 64K context on a 24GB RTX 3090, spilling eight MoE expert layers to CPU freed enough VRAM to increase -b from 512 to 1024 and -ub from 128 to 512. Prompt

b10295

model-releasesllama-cpp-releases

model-loader : fix quantized reshaped tensor strides (#26672) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64)

Best llama cpp flags to run Deepseek-flash 0731

Hi all. These are my system specs: dual xeon e5 2696 v2 , 160gb DDR3 ram ECC(1600mhz), 3 gpus: 3060 12gb, p100 16gb, 3050 6gb. And a 400gb nvme sdd RAID0, 3000 mb/s. The model is Deepseek-flash-0731 U

Safety

BridgeVLA++: A Data-Efficient, Generalizable, and Memory-Augmented Vision-Language-Action Framework for 3D Manipulation

arXiv:2608.05042v1 Announce Type: new Abstract: Leveraging pre-trained vision-language models (VLMs) to construct vision-language-action (VLA) models has emerged as a promising paradigm for 3D robot m

safetyarxiv-cs-ro

DeepSeek-V4 Flash 0731 vs GPT-5.6 Luna on DeepSWE: Cost and Coding

model-releasestogether-ai-blog

DeepSeek‑V4 Flash 0731 is the cheapest model on the DeepSWE board, costing about 0.10 per rollout versus GPT‑5.6 Luna’s 0.61, yet it scores a pass@1 of 53.3% compared to Luna’s 67.2%. A cascade strate

Research

Deltoris: Enabling Real-time VLA Inference in Embodied AI via Bit-level Sparsity and Speculative Inference

arXiv:2608.04428v1 Announce Type: cross Abstract: Vision-language-action (VLA) models have emerged as a key component in embodied AI. Among existing approaches, diffusion-based VLA models achieve supe

researcharxiv-cs-lg

EgoAfford: Task-Oriented Affordance Grounding via Egocentric Referring Segmentation

arXiv:2608.04533v1 Announce Type: new Abstract: Part-level affordance grounding has advanced the localization of functional object regions associated with elemental actions. Extending this capability

In addition to the upgrade in intelligence with GPT-5.6 Luna, Free and Go users can now use the “Think” button for more reasoning on harder …

OpenAI has released GPT‑5.6 Sol, which powers both instant and deep‑reasoning modes for ChatGPT Plus and Pro customers, delivering fact‑centric responses. Free and Go tier users will receive unlimited

model-releasesopenai--x