Search: “models”

MCIF: Multimodal Crosslingual Instruction-Following Benchmark from Scientific Talks

arXiv:2507.19634v4 Announce Type: replace-cross Abstract: Recent advances in large language models have laid the foundation for multimodal LLMs (MLLMs), which unify text, speech, and vision within a s

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

Rethinking Reasoning with MDLMs: Early Exits, Post-hoc Reasoning, and Beyond

DGX agent

arXiv:2510.19990v2 Announce Type: replace Abstract: The reasoning paradigm, where language models reason before answering, has enabled breakthroughs on tasks such as mathematical problem-solving. Whil

researcharxiv-cs-lg

11 Aug 2026

UNMASK: Discovering and Causally Verifying Spurious Shortcuts in Text Classifiers

arXiv:2608.09209v1 Announce Type: new Abstract: Neural language models trained on large crowdsourced corpora frequently exploit spurious surface patterns tied to target labels without true linguistic

VANE: Reliable Test-Time Training for Vision-Language-Action Models via Future Visual Representation Prediction

arXiv:2608.09448v1 Announce Type: cross Abstract: Test-time training (TTT) offers a lightweight way to adapt vision--language--action (VLA) policies from unlabeled deployment streams, but it remains d

safetyarxiv-cs-cv

VeinCast: Physics-Guided Dynamic Field Graphs with Graph-Conditioned Fusion for Global Medium-Range Weather Forecasting

arXiv:2608.09286v1 Announce Type: cross Abstract: Global medium-range weather forecasting requires modeling structured yet state-dependent interactions among heterogeneous atmospheric fields. Existing

When Grammar Guides the Attack: Uncovering Control-Plane Vulnerabilities in LLMs with Structured Output

arXiv:2503.24191v4 Announce Type: replace-cross Abstract: Content Warning: This paper may contain unsafe or harmful content generated by LLMs that may be offensive to readers. Large Language Models (L

Achievable 253 t/s - unsloth/Muse Glimmer 30B UD-Q5_K_M on a 5090

Benchmarked Muse Glimmer 30B on my RTX 5090 (32GB), 262k context, UD-Q5_K_M + dflash-kquant + mmproj. Workload Stock master + DFlash ngram-simple PR #26842 + DFlash Code patch 78 t/s 57 t/s 220-253 t/

An Exploratory Evaluation of LLM-Assisted Rewriting of Moderate-Complexity Financial Sentences for DisCoCat-Based Sentiment Analysis

arXiv:2608.07439v1 Announce Type: new Abstract: Quantum natural language processing (QNLP) provides a grammar-aware framework for text modeling, and Distributional Compositional Categorical (DisCoCat)

Are Visual Place Recognition Models Recognizing Places or Conditions? Distractor-Augmented Evaluation and Condition Suppression

arXiv:2608.06847v1 Announce Type: cross Abstract: Long-term Visual Place Recognition (VPR) is typically evaluated by matching queries from one condition against a database from another. Crowdsourced m

researcharxiv-cs-cv

Beyond 'AI Language': The case for the idiolectal nature of LLM output

arXiv:2608.06589v1 Announce Type: cross Abstract: While large language model outputs are frequently analysed as a collective super variety termed 'AI language,' this chapter argues that this perspecti

researcharxiv-cs-ai

CreativeInstruct: Scalably Teaching LLMs to Balance Quality, Creativity, and Diversity

arXiv:2608.07460v1 Announce Type: cross Abstract: While post-training improves the capabilities of large language models (LLMs), it generally lowers their output diversity and creativity, negatively i

safetyarxiv-cs-ai

DATAREEL: Automated Data-Driven Video Story Generation with Animations

arXiv:2604.25220v2 Announce Type: replace Abstract: Data videos combine animated visualizations with synchronized narration to communicate quantitative information and are widely used in journalism, e

Diffusion LLMs as Targets and Adversaries: Mechanistic Safety Exploits

arXiv:2608.07430v1 Announce Type: cross Abstract: Diffusion Large Language Models (DLLMs) replace autoregressive next-token prediction with iterative parallel denoising, yet their internal safety mech

Do AI Personas Grow? Analyzing and Benchmarking Personality Evolution in LLM Agents After Life Events

arXiv:2608.06485v1 Announce Type: cross Abstract: Personality-conditioned LLM agents (PC-Agents) are increasingly used in emotional support, social simulation, and role-playing, motivating the develop

From Optimal Actions to World Models: Identifiability of Transition Kernels in Discounted MDPs

arXiv:2608.07301v1 Announce Type: new Abstract: We study what can be recovered about the transition probabilities of a Markov decision process from optimal actions alone. This is closely related to th

safetyarxiv-cs-lg

Frontier performance you can actually own. Proud to help power DeepSeek-V4-Flash on Ollama's cloud, with the fastest hosted performance avai…

model-releasestogether-ai--x

Frontier performance you can actually own. Proud to help power DeepSeek-V4-Flash on Ollama's cloud, with the fastest hosted performance available. Open weights, zero data retention, U.S. & EU hosting.

Learning in Deep Networks under Dale's Constraint

arXiv:2608.06963v1 Announce Type: new Abstract: Biologically plausible learning models aim to explain how neural circuits can implement effective learning under the constraints of real neurons. Althou

Multi-Agent Forensic Reasoning for Generalizable Deepfake Video Detection

arXiv:2608.06865v1 Announce Type: cross Abstract: The malicious use of generative artificial intelligence to create highly realistic deepfake videos raises serious ethical concerns and poses substanti

Pre-Inference Routing for Cost-Efficient Document Field Extraction

arXiv:2608.06607v1 Announce Type: new Abstract: Most document-extraction systems use a single model for all documents. This is simple but can be costly for easy cases and less effective for difficult

researcharxiv-cs-cl

ReQuant: Fixed-Grid Discrete Refinement for Post-Training Quantization

arXiv:2608.07019v1 Announce Type: new Abstract: Post-training quantization (PTQ) is widely used to reduce the memory and computational cost of large language models. Existing PTQ methods typically obt

researcharxiv-cs-ai

Stoicheia: Character-Level Masked Diffusion for Ancient Greek Textual Restoration, Parsing, and Metrical Scansion

arXiv:2608.07249v1 Announce Type: new Abstract: We introduce Stoicheia, a 405M-parameter character-level masked-diffusion encoder for Ancient Greek whose input factors into five aligned, independently

Tensor Network Kernel Machines: A JAX Framework for Machine Learning and Nonlinear System Identification

model-releasesarxiv-cs-lg

arXiv:2608.07043v1 Announce Type: cross Abstract: Developing nonlinear models that are both expressive and computationally efficient remains a challenge in machine learning and nonlinear system identi

CyberKimi just dropped strong results on one of ExploitBench’s hardest V8 bugs , points away from Mythos

Hey everyone ! Quick share from the cyber + local LLM side of things that I found interesting. During this week’s hacker summer camp, an AI researcher and reverse malware engineer veteran 'lordx64' on

model-releasesr-ollama

9 Aug 2026

KLQ: Training-free measured rotation quantization. Beats all training-free rotation-based quantization methods on W4A4KV4-bits. Llama 3.2 1B KLQ-quantized beats SpinQuant and gets close to ReSpinQuant without GPTQ/LDLQ rounding.

First of all, I'm not a lab, this was a solo summer research project that finally culminated into the github repo and the writeup. The repo includes a much deeper dive with methods, findings about qua

9 Aug 2026

~45% lower MiniMax H3 sampler time with new Spectrum settings — degree 1 works surprisingly well (v0.1.8)

model-releasesr-stablediffusion

Follow-up to my original Spectrum MiniMax H3 post: https://www.reddit.com/r/StableDiffusion/comments/1vf1ze3/spectrum_acceleration_for_minimax_h3_in_comfyui/ In that first post, I released the MiniMax

Am I just hallucinating

Or is there any reason why I feel like model output quality seems to be better when I use higher micro-batch values (ub) in llama-cpp? I don't really have any hard numbers or anything (just running th

Arbitrage: Efficient Reasoning via Advantage-Aware Speculation

researchapple-ml-research

Modern Large Language Models achieve impressive reasoning capabilities with long Chain of Thoughts, but they incur substantial computational cost during inference, and this motivates techniques to imp

Basically every remaining good AI benchmark score has an implied asterisk next to it which reads: * could be signficantly higher with a bett…

model-releasesethan-mollick--x

On August 7, 2026 Ethan Mollick tweeted that “every remaining good AI benchmark score has an implied asterisk next to it which reads: * could be significantly higher with a better harness.” The commen

BendTwin: Robust Dense-to-Sparse Physical Reconstruction with Bending-Aware Differentiable Spring-Mass Models

arXiv:2608.06164v1 Announce Type: new Abstract: Reconstructing objects with mechanical properties from video observations enables physically consistent dynamic prediction, benefiting robotics planning

researcharxiv-cs-cv

Decoupling Perception from Description: Computation-Grounded Representation Alignment between Multivariate Time Series and Language

model-releasesarxiv-cs-lg

arXiv:2608.05238v1 Announce Type: new Abstract: Training multimodal models to align time series with language runs into a self-supervision trap. The usual recipe asks an LLM to read a series and write

Evaluating and Improving Pedagogical Fit in LLM-Based AI Tutors with the Pedagogical Suitability Index

arXiv:2608.05411v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used as AI tutors, but a correct answer is not always a pedagogically appropriate one. In classroom learni

FormBharo: Designing and Evaluating a Voice Agent for Conversational Form Filling in Rural India

arXiv:2608.06027v1 Announce Type: cross Abstract: In India, almost every social benefit starts with a form, yet the people who need these benefits most are often unable to read or write. Reaching them

Hijacking Robots with a Piece of Paper: A Systematic Study of Physical Prompt Injection in VLM-Controlled Robots

arXiv:2608.05715v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) are increasingly deployed as planners in robotic systems, where they translate natural-language commands into executable

Kastor: An efficient fine-tuning strategy for generative emulation of PDE simulations

model-releasesarxiv-cs-lg

arXiv:2608.06107v1 Announce Type: new Abstract: Machine learning offers a promising avenue to accelerate physical simulations by replacing computationally expensive traditional Partial Differential Eq

LangChoiceBench: Measuring and Explaining Programming-Language Choice in LLMs

arXiv:2608.06041v1 Announce Type: cross Abstract: Large language models (LLMs) have been shown to exhibit strong Python preferences when generating project-level code, but there is currently no system

Look Twice: Training-Free Evidence Highlighting for Knowledge-based Visual Question Answering

arXiv:2604.01280v2 Announce Type: replace-cross Abstract: Knowledge-based Visual Question Answering (KB-VQA) requires Multimodal Large Language Models (MLLMs) to identify and combine fine-grained visu

Tutorials

Measuring and Detecting Harmful AI Sycophancy

arXiv:2608.05624v1 Announce Type: new Abstract: Sycophantic responses are becoming pervasive in large language models (LLMs), and prior work has pointed out that some of them could be harmful. This pa

tutorialsarxiv-cs-ai

On-Policy Self-Distillation without Any Supervision

arXiv:2608.06296v1 Announce Type: new Abstract: On-policy (Self-)Distillation (OPD / OPSD) has shown strong potential for post-training large language models (LLMs). However, existing methods still re

safetyarxiv-cs-lg

Agents

OPERA: Operator-residual feedback for reliable autonomous optical experiments with language-model agents

arXiv:2608.05990v1 Announce Type: new Abstract: Autonomous agents choose actions using scores that may not reflect experimental success. We developed OPERA, an operator-residual framework for optical

agentsarxiv-cs-ai

Scaling Categorical Flow Maps

researchapple-ml-research

Continuous diffusion and flow matching models could represent a powerful alternative to autoregressive approaches for language modelling (LM), as they unlock a host of advantages currently reserved fo

Seeing Is Not Deciding: Can Multimodal LLMs Act as Effective CEOs?

arXiv:2608.05864v1 Announce Type: new Abstract: Large language models are increasingly applied as autonomous decision-making agents. However, in executive business decisions, existing benchmarks are l

Serving Deepseek v4 Flash 0731 on 2x DGX Spark — 5-7 GB OS headroom, what would you do to lower VRAM usage and increase OS available RAM?

Hey all, I'm serving DSv4Flash 0731 on a cluster of 2x DGX Sparks but am running into constant issues with having almost no RAM (unified memory) left for the OS/cache and I'd love to hear the communit

Shrinking the Generation-Verification Gap with Weak Verifiers

arXiv:2506.18203v3 Announce Type: replace Abstract: Verifiers can improve language model capabilities by scoring and ranking responses from generated candidates. Currently, high-quality verifiers are

Tutorials

Spectral Distillation: From Nonlinear Dynamics to Linear State-Space Models

arXiv:2608.05416v1 Announce Type: new Abstract: Can nonlinear dynamical systems be learned through a compact linear state-space representation, without directly solving a non-convex system-identificat

tutorialsarxiv-cs-lg

Subliminal Learning is Non-Semantic Distillation

arXiv:2608.05734v1 Announce Type: new Abstract: Subliminal Learning (SL) is a surprising type of generalization displayed by modern language models. It allows the transfer of a bias or behavior from a

Thanks for the thorough testing! With Qwen3.8-Max, everyone can observe the world in detail. 👀

Alibaba’s Qwen team announced the new Qwen‑3.8‑Max model and thanked users for extensive testing, noting its ability to provide highly detailed world observations. A community member highlighted that

model-releasesqwen--x

Why the Third Axis Is Freedom

arXiv:2608.05423v1 Announce Type: cross Abstract: In generative training, a model produces an output and is penalised for its difference from an example. With one output per comparison, a model that p

researcharxiv-cs-ai

Attention, Anomalies! Handling Attention Layers in Unsupervised Federated Outlier Detection