ninjasaid13

• Better & Faster Large Language Models via Multi-token Prediction

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 37 comments

• LongCat-Next: Lexicalizing Modalities as Discrete Tokens

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 4 comments

• View-oriented Conversation Compiler for Agent Trace Analysis

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• Intern-S1-Pro - 1 Trillion parameters Open-Weights for Scientific Multimodal Foundation Model

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 1 comments

• microsoft/Phi-4-reasoning-vision-15B · Hugging Face

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 1 comments

• LightMem: Lightweight and Efficient Memory-Augmented Generation

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 3 comments

• Can anyone tell me the performance of LLaVA vs BLIP?

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 2 comments

• OpenCUA: Open Foundations for Computer-Use Agents

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 10 comments

• Ming-flash-omni-Preview

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• Latent Sketchpad: Sketching Visual Thoughts to Elicit Multimodal Reasoning in MLLMs

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• Nvidia's OmniVinci: Enhancing Architecture and Data for Omni-Modal Understanding LLM

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 5 comments

• InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 13 comments

• [2510.13804] Generative Universal Verifier as Multimodal Meta-Reasoner

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• AndesVL Technical Report: An Efficient Mobile-side Multimodal Large Language Model

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• LLM-JEPA: Large Language Models Meet Joint Embedding Predictive Architectures

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 2 comments

• SDQ-LLM: Sigma-Delta Quantization for 1-bit LLMs of any size

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 2 comments

• An Open-source Omni Chatbot for Long Speech and Voice Clone

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 17 comments

• RLP: Reinforcement as a Pretraining Objective

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 1 comments

• Dynalang code released

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 2 comments

• NCSOFT/VARCO-VISION-2.0-14B · Hugging Face

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 13 comments

• Mini-o3: Scaling Up Reasoning Patterns and Interaction Turns for Visual Search

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• Kwai Keye-VL 1.5 Technical Report

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 4 comments

• Thyme: Think Beyond Images

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 1 comments

• Technical Report of TeleChat2, TeleChat2.5 and T1

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 3 comments

• MemOS: A Memory OS for AI System

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 19 comments

• Beyond Context Limits: Subconscious Threads for Long-Horizon Reasoning

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• Phi-4-mini-flash-reasoning

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 15 comments

• Code for Skywork-R1V3-38B

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• Kwai-Keye/Keye-VL-8B-Preview · Hugging Face

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 2 comments

• Stream-Omni: Simultaneous Multimodal Interactions with Large Language-Vision-Speech Model

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 5 comments

• inclusionAI/Ming-Lite-Omni · Hugging Face

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 11 comments

• Omni-R1: Reinforcement Learning for Omnimodal Reasoning via Two-System Collaboration

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 3 comments

• GitHub - jacklishufan/LaViDa: Official Implementation of LaViDa: :A Large Diffusion Language Model for Multimodal Understanding

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 3 comments

• Pretraining on the Test Set Is All You Need

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 2 comments

• Open-Sourced Multimodal Large Diffusion Language Models

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 17 comments

• GoT-R1: Unleashing Reasoning Capability of MLLM for Visual Generation with Reinforcement Learning

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 2 comments

• Qwen-14B model

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 1 comments

• Aya Vision: Advancing the Frontier of Multilingual Multimodality

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 17 comments

• R1-Reward: Training Multimodal Reward Model Through Stable Reinforcement Learning

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 2 comments

• Llama Nemotron - a nvidia Collection

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• DFloat11: Lossless LLM Compression for Efficient GPU Inference

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 6 comments

• Yo'Chameleon: Personalized Vision and Language Generation

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 1 comments

• Skywork-R1V2-38B - New SOTA open-source multimodal reasoning model

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 18 comments

• OpenGVLab/InternVL3-78B · Hugging Face

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 8 comments

• Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 1 comments

• Tina: Tiny Reasoning Models via LoRA

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 1 comments

• Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 0 comments

• Meta Perception Language Model: Enhancing Understanding of Visual Perception Tasks

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 27 comments

• An Easy-to-use Knowledge Editing Framework for LLMs.

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 4 comments

• Browser Qwen

  Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 16 comments