Search: “models”

A Measurement Study of LLM Inference Trade-offs Across Edge Continuum Hardware

arXiv:2609.08307v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used as backends for intelligent web services, but serving them across the edge continuum requires balan

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

AhaBench: Do Agents Learn from Prior Experience? A Benchmark for Long-Horizon Continual Learning

DGX agent

arXiv:2609.05435v1 Announce Type: new Abstract: Modern language agents are expected to operate over long horizons: they ask follow-up questions, reuse worked examples, handle tool feedback, and adapt

model-releasesarxiv-cs-lg

9 Sept 2026

All for 1-Bit: Towards Genuine 1-Bit Post-Training Quantization for LLMs

arXiv:2609.06161v1 Announce Type: new Abstract: Large language models (LLMs) have achieved remarkable progress, yet their massive storage and memory-bandwidth demands still hinder efficient deployment

Are Image Generators Zero-Shot Perceivers? A Rigorous Evaluation

arXiv:2609.07884v1 Announce Type: new Abstract: Recent work, such as Vision Banana, shows that lightweight instruction tuning can enable an image generator to achieve state-of-the-art performance acro

Safety

BAFF: Bid-Aware Filter Family for Mitigating Training Data Interference in RTB A/B Tests

arXiv:2609.08725v1 Announce Type: new Abstract: In online A/B tests for real-time bidding (RTB), control and treatment models are typically trained on a shared serving log that includes data generated

safetyarxiv-cs-lg

Benchmarking Storage Systems for Machine Learning Workloads Using NIO Bench

arXiv:2609.05418v1 Announce Type: cross Abstract: Machine learning training workloads place unique demands on storage systems, yet most existing benchmarks focus on computational throughput rather tha

CDFlow: Building Invertible Layers with Circulant and Diagonal Matrices

arXiv:2510.25323v4 Announce Type: replace Abstract: Normalizing flows are deep generative models that enable efficient likelihood estimation and sampling through invertible transformations. A key chal

CoSA: Correlation-Guided Change A ttention with Learnable Residual Gating for Remote Sensing Change Detection

arXiv:2609.08914v1 Announce Type: new Abstract: Pixel-level annotation of fixed traffic-camera imagery is expensive, while crosswalk models trained from street-level imagery face a substantial viewpoi

CrowdTraj: A Benchmark for Dense Crowd Trajectory Prediction in Realistic Crowded Environments

arXiv:2609.07685v1 Announce Type: new Abstract: In real-world applications, pedestrian trajectory prediction models rely on inputs from detection and tracking systems. Prior trajectory prediction benc

Disparity Has a Sign: Stereo Matching Beyond the Zero-Disparity Plane

arXiv:2609.06809v1 Announce Type: new Abstract: Modern stereo matching models fail when disparity crosses zero, with end-point error (EPE) rising by 4.6-37imes. Yet stereoscopic content, from cinema 3

Don't let FOMO win if you're interested in local llm from a hobby/learning aspect

Just a reminder for those out there itching to get into local llms - don't let FOMO or 'gear acquisition syndrom' take over. No matter the hobby, it's so easy to get stuck in a trap where we buy more

DriveZero: End-to-End Driving Beyond Human Demonstrations

arXiv:2609.06055v1 Announce Type: new Abstract: Most end-to-end autonomous-driving systems learn by imitating human driving logs, leaving their learned behavior constrained by the quality and behavior

Evolution of Multimodal Question Answering: From Modality-Adaptive Extraction to Unified Language Representation

arXiv:2609.08896v1 Announce Type: cross Abstract: The rapid growth of multimodal data has intensified the need for question answering (QA) systems capable of reasoning across heterogeneous sources suc

FANS: Federated Adaptive Network Search Learning for Heterogeneous Devices

arXiv:2609.06106v1 Announce Type: new Abstract: Heterogeneous Federated Learning (HFL) aims to train models across devices with diverse resource budgets while preserving data privacy. Existing HFL met

Forecasting the Winner of a Live Tennis Match

arXiv:2609.07617v1 Announce Type: new Abstract: With the rise of live sports betting in recent years, tennis forecasting has expanded from pre-match prediction to models that update win probabilities

researcharxiv-cs-lg

GLM 5.3 Flash Q4 @ 60tps / 550tps on M3 Ultra

I have been a dwarfstar fan for awhile and I really liked glm 5.3 flash but needed it to be materially faster to feel good using it. In the screenshot you can see the outcome of using the model with a

GraphNOSE: A Graph Transformer in Olfaction

arXiv:2609.05694v1 Announce Type: new Abstract: Predicting olfactory qualities from molecular structure is an open problem in chemoinformatics. Although linear models can link molecular features to od

researcharxiv-cs-lg

Hidden in Plain Sight: The Overlooked Significance of Canonical Elements for Extreme LLM Sparsity

arXiv:2609.06557v1 Announce Type: new Abstract: Large language models (LLMs) are often considered fragile under aggressive sparsification, and maintaining reliable performance typically requires stick

Human-Centric Image Captioning with Subject-Centered Spatial Understanding

arXiv:2609.08300v1 Announce Type: new Abstract: While multimodal large language models (MLLMs) achieve remarkable performance on generic image captioning, they frequently suffer from structural halluc

ICYMI: What landed for AI builders in August 2026

model-releasesaws-ml-blog

A recap of August 2026 launches for AI builders across Amazon Bedrock, Amazon Bedrock AgentCore, and Strands: million-token context for OpenAI models, cross-Region inference, agents that run for up to

Safety

If they truly believe the risk is that high, and I think they do, then we need 100x more research and understanding of it. That requires the…

If they truly believe the risk is that high, and I think they do, then we need 100x more research and understanding of it. That requires them to openly share their models, datasets, training code, and

safetyclem-delangue--x

Inducing Emergent Misalignment from Reward Hacks with Iterative DPO

arXiv:2609.06649v1 Announce Type: new Abstract: Reward hacking during reinforcement learning from verifiable rewards (RLVR) can induce reward seeking and broad misalignment in language models. Studyin

IXPLORE: Bounded Ideal Point Estimation with Grid-Based Uncertainty Quantification

arXiv:2609.06018v1 Announce Type: new Abstract: Ideal point estimation is widely used to analyze and visualize political data. However, selecting the corresponding spatial model involves various trade

LatentFM: A Latent Flow Matching Approach for Generative Medical Image Segmentation

arXiv:2512.04821v4 Announce Type: replace Abstract: Generative models have achieved remarkable progress with the emergence of flow matching (FM). It has demonstrated strong generative capabilities and

researcharxiv-cs-cv

Safety

Let It Go or Learn to Self-Correct: Continuous Diffusion for Constrained Discrete Tasks

arXiv:2609.09009v1 Announce Type: new Abstract: Denoising Diffusion Probabilistic Models (DDPMs) generate samples by starting from noise and repeatedly denoising while keeping each update close to the

safetyarxiv-cs-lg

One Rate Is Not Enough: Adaptive Anisotropic Learning Rates for LoRA Fine-Tuning

arXiv:2609.05885v1 Announce Type: new Abstract: Low-rank adaptation (LoRA) has become the standard for parameter-efficient fine-tuning of large language models. Most LoRA variants follow a uniform-LR

Remove360: Benchmarking Residuals After Object Removal in 3D Gaussian Splatting

arXiv:2508.11431v3 Announce Type: replace Abstract: An object can disappear from a 3D scene, yet still be detectable. Even after visual removal, modern vision models may infer what was originally pres

Robust Vision Systems for Connected and Autonomous Vehicles: Security Challenges and Attack Vectors

arXiv:2602.09740v3 Announce Type: replace Abstract: Connected and Autonomous Vehicles (CAVs), which is critical for achieving Level-5 autonomous driving. Safe and reliable CAV navigation depends on ro

Steering Geometry: Validating Human Value Geometry in LLM Steering Space

arXiv:2609.06289v1 Announce Type: cross Abstract: As large language models (LLMs) are increasingly deployed in alignment-sensitive contexts, activation steering has emerged as a lightweight, inference

Steering Under Compression: Dose-Response, Capability Cost, and Failure Asymmetry in Quantized LLMs

arXiv:2609.06473v1 Announce Type: new Abstract: Inference-time activation steering enables behavioral control of large language models without parameter modification, while post-training quantization

Local Ai

v0.34.0

Use Ollama models in ChatGPT Desktop Ollama models can now be used directly in ChatGPT Desktop, so you can keep your existing workflow while running open models. Setup is available from the Ollama app

local-aiollama-releases

What the Window Does Not Contain: Auditing Provenance in a Document-Grounded Instability Benchmark

arXiv:2609.06147v1 Announce Type: cross Abstract: Ask a language model the same question about the same document twenty times, and it sometimes returns two different answers. We built Probity, a bench

You Can't Prefer Emotions You Don't Sample: Intensity Undershoot in DPO-Tuned LLMs

arXiv:2609.07808v1 Announce Type: cross Abstract: Ask a language model to respond 'very excitedly,' and its output is typically only mildly more energetic. We quantify this effect. We condition an ins

Generating Bad Apple autonomously from a single initial state using a tiny recurrent dynamical system (417k params) [P]

model-releasesr-machinelearning

A few weeks ago, I saw this post where the author trained a SIREN MLP to implicitly memorize Bad Apple as a coordinate function: (t, y, x) to pixel. That got me curious about a slightly different form

I reduced image-processing token usage by ~95% compared with GPT-4o direct vision, while maintaining roughly the same accuracy.How significant is that?[P]

model-releasesr-machinelearning

I'm testing a new approach for reducing the cost of image-based LLM inference. I evaluated it on the MOMA Graph benchmark, using 1,315 questions. Compared with using GPT-4o to process the original ima

inclusionAI/Ling-3.0-flash-VL · Hugging Face

Ling-3.0-flash-VL inherits the language, reasoning, and long-context capabilities of Ling-3.0-flash, while extending them with native image and video understanding. The model has 124B total parameters

Qwen3.8-Flash-Next in llama.cpp vs SGLang vs FreeToken: 35s vs 258s to first token at full context. My findings on new PRs coming to engines.

Hey guys, After my CPU-only to 96GB VRAM test, I tested Qwen3.8-Flash-Next across llama.cpp, SGLang and FreeToken on the same workstation. This time I wanted to see what changes when you keep the hard

A Calibrated Reflection Approach for Enhancing Confidence Estimation in LLMs

model-releasesarxiv-cs-cl

arXiv:2609.04539v1 Announce Type: new Abstract: A critical challenge in deploying Large Language Models (LLMs) is developing reliable mechanisms to estimate their confidence, enabling systems to deter

DeepSeek-V4-Flash-Vision-Exp is amazing at creating game worlds!

Model: DeepSeek-V4-Flash-Vision-Exp (local and API when impatient) Time: about one weekend (2 days) of QA and small improvements Full game is here After Qwen3.8-Flash-Next one-shotted a really cool Ca

// 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

Extremely Sparse Supervision Incentivizes Reasoning Ability

arXiv:2609.04565v1 Announce Type: new Abstract: Large language models demonstrate increasingly strong reasoning capabilities through effective post-training. Yet, prevailing post-training methods opti

Krea2 Turbo Distill 4 step LoRA - FINAL Version released today

model-releasesr-stablediffusion

Krea 2 Turbo — 4-Step Distillation LoRA Full details and to download - HF Repo: https://huggingface.co/lvladikov/Krea2-Turbo-Distill-4step-LoRA https://preview.redd.it/9p2rjwh3y3oh1.jpg?width=3072&for

PRICE: A Systematic Study of LLM Adaptation Choices for Bitcoin Price Forecasting

arXiv:2609.05235v1 Announce Type: cross Abstract: Cryptocurrency markets exhibit extreme volatility and non-stationary dynamics that challenge conventional forecasting methods. Although Large Language

RISE: Recursive Improvement via Self-Extrapolating Policy Distillation

arXiv:2609.05295v1 Announce Type: new Abstract: On-policy distillation (OPD) provides dense, per-token supervision for language model post-training, but its effectiveness is bottlenecked by teacher qu

Tracing Audio Grounding and Answer Selection in Audio LLMs

arXiv:2609.04637v1 Announce Type: cross Abstract: Audio Large Language Models (Audio LLMs) have advanced in audio understanding, yet they can still predict the answer by reasoning from textual cues or

researcharxiv-cs-ai

WeAgent-MMGenEdit: A Full-Stack Recipe for Multimodal Agentic Image Generation and Editing

arXiv:2609.05171v1 Announce Type: new Abstract: Image generation and editing models have advanced rapidly, yet remain unreliable when prompts require external world knowledge. Bounded and long-tail pa

When Seeing Overrides Knowing: Visual Dominance and Deferral-Based Method for Personalized Safety in VLMs

arXiv:2609.04281v1 Announce Type: cross Abstract: Vision-language models (VLMs) are increasingly deployed in high-stakes settings, where a response that is reasonable in general may still be unsafe fo