model releases

PersonaVLM: Long-Term Personalized Multimodal LLMs

arXiv:2604.13074v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) serve as daily assistants for millions. However, their ability to generate responses aligned with individual pr

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Predicting Time Pressure of Powered Two-Wheeler Riders for Proactive Safety Interventions

DGX agent

arXiv:2601.03173v2 Announce Type: replace Abstract: Time pressure critically influences risky maneuvers and crash proneness among powered two-wheeler riders, yet its prediction remains underexplored i

model-releasesarxiv-cs-lg

16 Apr 2026

Qwen 3.6 is here, and open-source! Run it locally with improved agentic coding capabilities. Try it with Claude Code: ollama launch claude -…

Qwen 3.6 is here, and open-source! Run it locally with improved agentic coding capabilities. Try it with Claude Code: ollama launch claude --model qwen3.6 Try it with OpenClaw: ollama launch openclaw

model-releasesollama--x

Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7

model-releasessimon-willison

For anyone who has been (inadvisably) taking my pelican riding a bicycle benchmark seriously as a robust way to test models, here are pelicans from this morning's two big model releases - Qwen3.6-35B-

RAG or Learning? Understanding the Limits of LLM Adaptation under Continuous Knowledge Drift in the Real World

arXiv:2604.05096v2 Announce Type: replace Abstract: Large language models (LLMs) acquire most of their knowledge during pretraining, which ties them to a fixed snapshot of the world and makes adaptati

RANDPOL: Parameter-Efficient End-to-End Quadruped Locomotion via Randomized Policy Learning

arXiv:2505.19054v2 Announce Type: replace Abstract: Modern learning-based locomotion controllers typically rely on fully trainable deep neural networks with a large number of parameters. This paper st

Rare Event Analysis via Stochastic Optimal Control

arXiv:2604.13213v1 Announce Type: cross Abstract: Rare events such as conformational changes in biomolecules, phase transitions, and chemical reactions are central to the behavior of many physical sys

ReConText3D: Replay-based Continual Text-to-3D Generation

arXiv:2604.13730v1 Announce Type: new Abstract: Continual learning enables models to acquire new knowledge over time while retaining previously learned capabilities. However, its application to text-t

Red Skills or Blue Skills? A Dive Into Skills Published on ClawHub

arXiv:2604.13064v1 Announce Type: new Abstract: Skill ecosystems have emerged as an increasingly important layer in Large Language Model (LLM) agent systems, enabling reusable task packaging, public d

Replit Agent 4 is even smarter now with Claude Opus 4.7! 50% off for a limited time. Go try it now ↓

Replit Agent 4 has been upgraded to use Claude Opus 4.7, an advanced AI model, enhancing its code generation and problem-solving capabilities. The company is offering a 50% discount for a limited time

model-releasesreplit--x

Response.

model-releasesethan-mollick--x

Response. Hey Ethan! Sean here, PM on http://Claude.ai - thanks for the feedback. This isn't a router, this is the model being trained to decide when to think based on the context -- we've been runnin

Reward Design for Physical Reasoning in Vision-Language Models

arXiv:2604.13993v1 Announce Type: cross Abstract: Physical reasoning over visual inputs demands tight integration of visual perception, domain knowledge, and multi-step symbolic inference. Yet even st

Reward Hacking in the Era of Large Models: Mechanisms, Emergent Misalignment, Challenges

arXiv:2604.13602v1 Announce Type: new Abstract: Reinforcement Learning from Human Feedback (RLHF) and related alignment paradigms have become central to steering large language models (LLMs) and multi

RiskWebWorld: A Realistic Interactive Benchmark for GUI Agents in E-commerce Risk Management

arXiv:2604.13531v1 Announce Type: cross Abstract: Graphical User Interface (GUI) agents show strong capabilities for automating web tasks, but existing interactive benchmarks primarily target benign,

ROBOGATE: Adaptive Failure Discovery for Safe Robot Policy Deployment via Two-Stage Boundary-Focused Sampling

model-releasesarxiv-cs-ro

arXiv:2603.22126v3 Announce Type: replace Abstract: Deploying learned robot manipulation policies in industrial settings requires rigorous pre-deployment validation, yet exhaustive testing across high

Robust Low-Rank Tensor Completion based on M-product with Weighted Correlated Total Variation and Sparse Regularization

arXiv:2604.13525v1 Announce Type: cross Abstract: The robust low-rank tensor completion problem addresses the challenge of recovering corrupted high-dimensional tensor data with missing entries, outli

Robust Reward Modeling for Large Language Models via Causal Decomposition

arXiv:2604.13833v1 Announce Type: new Abstract: Reward models are central to aligning large language models, yet they often overfit to spurious cues such as response length and overly agreeable tone.

ROSE: Retrieval-Oriented Segmentation Enhancement

arXiv:2604.14147v1 Announce Type: new Abstract: Existing segmentation models based on multimodal large language models (MLLMs), such as LISA, often struggle with novel or emerging entities due to thei

RPS: Information Elicitation with Reinforcement Prompt Selection

arXiv:2604.13817v1 Announce Type: new Abstract: Large language models (LLMs) have shown remarkable capabilities in dialogue generation and reasoning, yet their effectiveness in eliciting user-known bu

Scaling Test-Time Compute to Achieve IOI Gold Medal with Open-Weight Models

arXiv:2510.14232v2 Announce Type: replace-cross Abstract: Competitive programming has become a rigorous benchmark for evaluating the reasoning and problem-solving capabilities of large language models

Seedance 2.0: Advancing Video Generation for World Complexity

arXiv:2604.14148v1 Announce Type: new Abstract: Seedance 2.0 is a new native multi-modal audio-video generation model, officially released in China in early February 2026. Compared with its predecesso

Seek-and-Solve: Benchmarking MLLMs for Visual Clue-Driven Reasoning in Daily Scenarios

arXiv:2604.14041v1 Announce Type: new Abstract: Daily scenarios are characterized by visual richness, requiring Multimodal Large Language Models (MLLMs) to filter noise and identify decisive visual cl

Self-Organizing Maps with Optimized Latent Positions

arXiv:2604.13622v1 Announce Type: new Abstract: Self-Organizing Maps (SOM) are a classical method for unsupervised learning, vector quantization, and topographic mapping of high-dimensional data. Howe

SemAttNet: Towards Attention-based Semantic Aware Guided Depth Completion

arXiv:2204.13635v2 Announce Type: replace Abstract: Depth completion involves recovering a dense depth map from a sparse map and an RGB image. Recent approaches focus on utilizing color images as guid

SHARe-KAN: Post-Training Vector Quantization for Cache-Resident KAN Inference

arXiv:2512.15742v2 Announce Type: replace Abstract: Pre-trained Vision Kolmogorov-Arnold Networks (KANs) store a dense B-spline grid on every edge, inflating prediction-head parameter counts by more t

Shocking result on my pelican benchmark this morning, I got a better pelican from a 21GB local Qwen3.6-35B-A3B running on my laptop than I d…

model-releasesclem-delangue--x

Shocking result on my pelican benchmark this morning, I got a better pelican from a 21GB local Qwen3.6-35B-A3B running on my laptop than I did from the new Opus 4.7! Qwen on the left, Opus on the righ

SLQ: Bridging Modalities via Shared Latent Queries for Retrieval with Frozen MLLMs

arXiv:2604.13710v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) exhibit strong reasoning and world knowledge, yet adapting them for retrieval remains challenging. Existing app

SocialMirror: Reconstructing 3D Human Interaction Behaviors from Monocular Videos with Semantic and Geometric Guidance

arXiv:2604.13581v1 Announce Type: new Abstract: Accurately reconstructing human behavior in close-interaction scenarios is crucial for enabling realistic virtual interactions in augmented reality, pre

Sorry for the long wait, everyone! As I said, Qwen is going to keep open-sourcing!

model-releasesclem-delangue--x

Sorry for the long wait, everyone! As I said, Qwen is going to keep open-sourcing! ⚡ Meet Qwen3.6-35B-A3B：Now Open-Source！🚀🚀 A sparse MoE model, 35B total params, 3B active. Apache 2.0 license. 🔥 Agen

SparseBalance: Load-Balanced Long Context Training with Dynamic Sparse Attention

arXiv:2604.13847v1 Announce Type: new Abstract: While sparse attention mitigates the computational bottleneck of long-context LLM training, its distributed training process exhibits extreme heterogene

SpatialEvo: Self-Evolving Spatial Intelligence via Deterministic Geometric Environments

arXiv:2604.14144v1 Announce Type: cross Abstract: Spatial reasoning over three-dimensional scenes is a core capability for embodied intelligence, yet continuous model improvement remains bottlenecked

Spectral Entropy Collapse as an Empirical Signature of Delayed Generalisation in Grokking

arXiv:2604.13123v1 Announce Type: new Abstract: Grokking -- delayed generalisation long after memorisation -- lacks a predictive mechanistic explanation. We identify the normalised spectral entropy il

Stein Variational Uncertainty-Adaptive Model Predictive Control

model-releasesarxiv-cs-ro

arXiv:2604.01034v2 Announce Type: replace Abstract: We propose a Stein variational distributionally robust controller for nonlinear dynamical systems with latent parametric uncertainty. The method is

Stochastic Trust-Region Methods for Over-parameterized Models

arXiv:2604.14017v1 Announce Type: cross Abstract: Under interpolation-type assumptions such as the strong growth condition, stochastic optimization methods can attain convergence rates comparable to f

Syn-TurnTurk: A Synthetic Dataset for Turn-Taking Prediction in Turkish Dialogues

arXiv:2604.13620v1 Announce Type: new Abstract: Managing natural dialogue timing is a significant challenge for voice-based chatbots. Most current systems usually rely on simple silence detection, whi

Synthesis and Deployment of Maximal Robust Control Barrier Functions through Adversarial Reinforcement Learning

model-releasesarxiv-cs-ro

arXiv:2604.13192v1 Announce Type: cross Abstract: Robust control barrier functions (CBFs) provide a principled mechanism for smooth safety enforcement under worst-case disturbances. However, existing

Synthesizing Instruction-Tuning Datasets with Contrastive Decoding

arXiv:2604.13538v1 Announce Type: new Abstract: Using responses generated by high-performing large language models (LLMs) for instruction tuning has become a widely adopted approach. However, the exis

Synthetic Tabular Generators Fail to Preserve Behavioral Fraud Patterns: A Benchmark on Temporal, Velocity, and Multi-Account Signals

arXiv:2604.13125v1 Announce Type: new Abstract: We introduce behavioral fidelity -- a third evaluation dimension for synthetic tabular data that measures whether generated data preserves the temporal,

Target-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?

arXiv:2511.17792v2 Announce Type: replace Abstract: While recent video world models can generate highly realistic videos, their ability to perform semantic reasoning and planning remains unclear and u

Text-as-Signal: Quantitative Semantic Scoring with Embeddings, Logprobs, and Noise Reduction

arXiv:2604.13056v1 Announce Type: new Abstract: This paper presents a practical pipeline for turning text corpora into quantitative semantic signals. Each news item is represented as a full-document e

Text-Attributed Knowledge Graph Enrichment with Large Language Models for Medical Concept Representation

arXiv:2604.13331v1 Announce Type: new Abstract: In electronic health record (EHR) mining, learning high-quality representations of medical concepts (e.g., standardized diagnosis, medication, and proce

The cognitive companion: a lightweight parallel monitoring architecture for detecting and recovering from reasoning degradation in LLM agents

arXiv:2604.13759v1 Announce Type: cross Abstract: Large language model (LLM) agents on multi-step tasks suffer reasoning degradation, looping, drift, stuck states, at rates up to 30% on hard tasks. Cu

The Consciousness Cluster: Emergent preferences of Models that Claim to be Conscious

arXiv:2604.13051v1 Announce Type: new Abstract: There is debate about whether LLMs can be conscious. We investigate a distinct question: if a model claims to be conscious, how does this affect its dow

The UK unveils Sovereign AI, a £500M fund to invest in domestic AI startups, starting with Callosum, which builds software to help different chips work together (Joel Khalili/Wired)

Joel Khalili / Wired: The UK unveils Sovereign AI, a £500M fund to invest in domestic AI startups, starting with Callosum, which builds software to help different chips work together — In a bid to min

model-releasestechmeme

These updates are rolling out on the Codex desktop app starting today. https://openai.com/index/codex-for-almost-everything/

OpenAI announced the rollout of updates to the Codex desktop application, beginning on the date of the announcement. The updates likely enhance Codex's code generation and AI-assisted programming capa

model-releasesopenai--x

TIP: Token Importance in On-Policy Distillation

arXiv:2604.14084v1 Announce Type: new Abstract: On-policy knowledge distillation (OPD) trains a student on its own rollouts under token-level supervision from a teacher. Not all token positions matter

TLoRA+: A Low-Rank Parameter-Efficient Fine-Tuning Method for Large Language Models

arXiv:2604.13368v1 Announce Type: new Abstract: Fine-tuning large language models (LLMs) aims to adapt pre-trained models to specific tasks using relatively small and domain-specific datasets. Among P

Today we’re announcing Ternary Bonsai: Top intelligence at 1.58 bits Using ternary weights {-1, 0, +1}, we built a family of models that are…

model-releasesemad-mostaque--x

Today we’re announcing Ternary Bonsai: Top intelligence at 1.58 bits Using ternary weights {-1, 0, +1}, we built a family of models that are 9x smaller than their 16-bit counterparts while outperformi