Papersby GlyphFoundation models
Technical reports for new model families. 59 papers from 21 labs.
Nemotron-Labs-Diffusion-Uni: Elastic Self-Refining Unified Diffusion LLM with Cross-Paradigm On-Policy DistillationNLD-Uni is a unified discrete diffusion language model that combines visual understanding and generation via cross-paradigm on-policy distillation and progressive self-refining diffusion.
NVIDIA12 upvotesRead →Eat Your Precision: Precision Objective Mitigates Reward Hacking in Rubric-based RLFORK jointly optimizes recall and precision via rubric-driven necessity credits and factual grading to improve response quality and reduce verbosity in reinforcement learning.
Carnegie Mellon University34 upvotesRead →Procedural Graphs: Self-Evolving Execution Structures for LLM AgentsA procedural graph framework organizes agent actions into structured relational triplets, providing situational guidance and self-evolving topology to improve long-horizon tool use.
Google2 citations41 upvotesRead →Instella-MoE Technical ReportInstella-MoE is an open Mixture-of-Experts language model trained on AMD GPUs using sparse activation, Gated Multi-head Latent Attention, and multi-stage post-training to achieve strong benchmark performance with full reproducibility.
AMD3 upvotesRead →DiffusionGemma Technical ReportDiffusionGemma is a fine-tuned mixture-of-experts language model that uses discrete diffusion to generate text blocks in parallel, achieving high speed while preserving capabilities like multimodal inputs and reasoning.
Google DeepMind14 citations40 upvotesRead →MerchantBench: Benchmarking LLM Agents for Long-Term Coherence in E-Commerce OperationsMerchantBench evaluates long-term coherence of LLM agents in a year-long e-commerce simulation requiring persistent decision-making across sourcing, pricing, and cash-flow management with mixed feedback delays.
Alibaba (Qwen)5 citations112 upvotesRead →Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI AgentsQwen-UI-Agent is a foundation GUI agent that integrates mobile, computer, web, and search environments with CLI execution, long-horizon online reinforcement learning, and autonomous data generation to achieve state-of-the-art real-world device performance.Alibaba (Qwen)18 citations156 upvotesRead →Qwen-Music Technical ReportQwen-Music is a music generation system that uses tokenized semantic representations, autoregressive language modeling with melody planning, and generative rendering to produce high-fidelity songs from text or reference audio.Alibaba (Qwen)4 citations32 upvotesRead →Gemma 4 Technical ReportOpen multimodal models from 2.3B to 31B parameters, with a thinking mode and image and audio input.Google DeepMind356 citations81 upvotesRead →Reinforcement Learning with Metacognitive Feedback Elicits Faithful Uncertainty Expression in LLMsReinforcement learning with metacognitive feedback and metacognitive data selection improve large language model calibration by enabling accurate self-assessment of performance and uncertainty.Google3 citations28 upvotesRead →Qwen-Image-2.0-RL Technical ReportA reinforcement learning and on-policy distillation approach enhances the visual quality and instruction-following capabilities of a diffusion model for image generation and editing tasks.Alibaba (Qwen)10 citations54 upvotesRead →AI Exposure Scores: what they measure, what they miss, and what comes nextReviews how widely cited scores of which jobs AI can assist are used in policy, what they miss, and newer measures that address it.
CohereRead →Qwen-RobotNav Technical Report: A Scalable Navigation Model Designed for an Agentic Navigation SystemQwen-RobotNav is a scalable navigation model with a parameterized interface enabling dynamic task modes and observation parameters, achieving state-of-the-art performance through multi-task training and zero-shot generalization to real-world robotics.Alibaba (Qwen)23 citations29 upvotesRead →Qwen-RobotManip Technical Report: Alignment Unlocks Scale for Robotic Manipulation Foundation ModelsA Vision-Language-Action foundation model for robotic manipulation achieves generalization through unified alignment across representation, motion, and behavior dimensions, enabling large-scale training on diverse data sources.Alibaba (Qwen)72 citations34 upvotesRead →Qwen-RobotWorld Technical Report: Unifying Embodied World Modeling through Language-Conditioned Video GenerationQwen-RobotWorld is a language-conditioned video world model that predicts future visual trajectories across multiple robotic domains using a double-stream diffusion transformer and embodied world knowledge corpus.Alibaba (Qwen)16 citations47 upvotesRead →From AGI to ASIArtificial general intelligence development may lead to artificial general superintelligence through multiple pathways, requiring interdisciplinary global preparation for transformative societal changes.Google9 citations39 upvotesRead →Soft-SVeRL: Self-Verified Reinforcement Learning with Soft RewardsTurns each prompt into a checklist scored item by item, giving reinforcement learning partial-credit rewards for tasks that cannot be checked automatically.Cohere1 citationsRead →Gated DeltaNet-2: Decoupling Erase and Write in Linear AttentionGated DeltaNet-2 improves upon existing linear attention models by separating erase and write operations through distinct channel-wise gates, achieving superior performance in long-context language modeling and retrieval tasks.NVIDIA48 citations30 upvotesRead →Qwen-Image-VAE-2.0 Technical ReportQwen-Image-VAE-2.0 is a high-compression Variational Autoencoder suite that improves reconstruction fidelity and diffusability through enhanced architecture, large-scale training, and semantic alignment strategies.Alibaba (Qwen)2 citations61 upvotesRead →Qwen-Image-2.0 Technical ReportQwen-Image-2.0 is an advanced image generation model that combines high-fidelity synthesis with precise editing capabilities through a unified framework using Qwen3-VL as condition encoder and Multimodal Diffusion Transformer for joint modeling.Alibaba (Qwen)40 citations118 upvotesRead →Seeing the Needle in the Haystack: Towards Weakly-Supervised Log Instance Anomaly Localization via Counterfactual PerturbationLogMILP is a weakly supervised framework for log anomaly detection that enables both bag-level detection and instance-level localization using prototype-guided structural modeling with counterfactual perturbation consistency regularization.
Hugging Face4 upvotesRead →Beyond the Assistant Turn: User Turn Generation as a Probe of Interaction Awareness in Language ModelsUser-turn generation serves as a probe to measure interaction awareness in large language models, revealing that this capability is distinct from task accuracy and can be influenced by training methods.
Salesforce5 upvotesRead →Embarrassingly Simple Self-Distillation Improves Code GenerationSimple self-distillation improves code generation in large language models by fine-tuning on model-generated samples, effectively addressing precision-exploration trade-offs in decoding.
Apple40 citations55 upvotesRead →Understanding the Challenges in Iterative Generative Optimization with LLMsGenerative optimization using large language models faces challenges due to implicit design decisions about artifact modification and learning evidence that significantly impact success across different applications.Google DeepMind11 citations27 upvotesRead →Multi-User Large Language Model AgentsMulti-user large language model agents face challenges in handling conflicting objectives, privacy preservation, and coordination efficiency in multi-principal decision-making scenarios.
Stanford University9 citations28 upvotesRead →VIBEVOICE-ASR Technical ReportVibeVoice-ASR is a unified end-to-end speech understanding framework that processes long-form audio without chunking, supports multiple languages and code-switching, and uses prompt-based context injection for improved domain-specific accuracy.
Microsoft31 citations25 upvotesRead →Fish Audio S2 Technical ReportFish Audio S2 is an open-source text-to-speech system with multi-speaker capabilities, multi-turn generation, and instruction-following control through natural-language descriptions, utilizing a multi-stage training approach and production-ready inference engine.
Fish Audio24 citations40 upvotesRead →Qwen3-Coder-Next Technical ReportQwen3-Coder-Next is an 80-billion-parameter language model that activates only 3 billion parameters during inference, achieving strong coding capabilities through agentic training with verifiable task synthesis and reinforcement learning.Alibaba (Qwen)95 citations66 upvotesRead →dLLM: Simple Diffusion Language ModelingA unified open-source framework is presented that standardizes core components of diffusion language modeling for reproduction, customization, and accessible development of both large and small models.
UC Berkeley18 citations136 upvotesRead →Arcee Trinity Large Technical ReportArcee Trinity models are sparse Mixture-of-Experts architectures with varying parameter counts and activation patterns, utilizing advanced attention mechanisms and training optimizations.
Arcee AI8 citations22 upvotesRead →Qwen3-ASR Technical ReportThe Qwen3-ASR family introduces speech recognition models with language identification capabilities and a non-autoregressive forced alignment model, achieving state-of-the-art performance and efficient processing.Alibaba (Qwen)200 citations43 upvotesRead →Qwen3-TTS Technical ReportThe Qwen3-TTS series presents advanced multilingual text-to-speech models with voice cloning and controllable speech generation capabilities, utilizing dual-track LM architecture and specialized speech tokenizers for efficient streaming synthesis.Alibaba (Qwen)194 citations80 upvotesRead →SimMerge: Learning to Select Merge Operators from Similarity SignalsPredicts which way of merging language models will work best from cheap similarity signals, avoiding costly merge-and-test searches.Cohere3 citationsRead →NitroGen: An Open Foundation Model for Generalist Gaming AgentsNitroGen is a vision-action foundation model trained on extensive gameplay data that demonstrates strong cross-game generalization and effective transfer learning capabilities.NVIDIA29 citations46 upvotesRead →Qwen3-VL Technical ReportQwen3-VL, a vision-language model, excels in text and multimodal understanding through advanced architectures and larger contexts, achieving superior performance across benchmarks.Alibaba (Qwen)3,178 citations164 upvotesRead →Tongyi DeepResearch Technical ReportTongyi DeepResearch, a large language model with agentic capabilities, achieves top performance in various deep research tasks through an end-to-end training framework and automatic data synthesis.Alibaba (Qwen)29 citations105 upvotesRead →Language Models Model LanguageThe paper advocates for an empiricist approach to evaluating language models, emphasizing frequency of use over traditional theoretical frameworks.
Snowflake26 upvotesRead →Making, not Taking, the Best of NFusion-of-N (FusioN) method improves LLM generation quality by synthesizing elements from multiple samples, outperforming Best-of-N in various settings and tasks.Cohere7 citations11 upvotesRead →TruthRL: Incentivizing Truthful LLMs via Reinforcement LearningTruthRL, a reinforcement learning framework, enhances the truthfulness of large language models by balancing accuracy and abstention, significantly reducing hallucinations and improving performance across benchmarks.
Meta24 citations56 upvotesRead →Qwen3-Omni Technical ReportQwen3-Omni, a multimodal model, achieves state-of-the-art performance across text, image, audio, and video, using a Thinker-Talker MoE architecture and a lightweight causal ConvNet for efficient streaming synthesis.Alibaba (Qwen)653 citations155 upvotesRead →VibeVoice Technical ReportVibeVoice synthesizes long-form multi-speaker speech using next-token diffusion and a highly efficient continuous speech tokenizer, achieving superior performance and fidelity.Microsoft56 citations180 upvotesRead →Qwen3 Technical ReportQwen3, a unified series of large language models, integrates thinking and non-thinking modes, reduces computational resources, and achieves state-of-the-art performance across various tasks and languages.Alibaba (Qwen)10.1k citations339 upvotesRead →LLM Defenses Are Not Robust to Multi-Turn Human Jailbreaks YetMulti-turn human jailbreaks reveal significant vulnerabilities in large language model defenses, achieving high attack success rates and exposing flaws in machine unlearning mechanisms.
Scale AI186 citationsRead →The Llama 3 Herd of ModelsOpen models up to 405B parameters that match leading closed models on many tasks.Meta20.8k citations119 upvotesRead →Nemotron-4 340B Technical ReportOpen models sized for one 8-GPU server, aligned with over 98% synthetic data.NVIDIA142 citationsRead →Replacing Judges with Juries: Evaluating LLM Generations with a Panel of Diverse ModelsEvaluating Large Language Models using a Panel of smaller evaluators outperforms single large evaluators with reduced bias and cost.Cohere418 citations71 upvotesRead →Phi-3 Technical Report: A Highly Capable Language Model Locally on Your PhoneSmall models trained on filtered and synthetic data that run on a phone and rival much larger ones.Microsoft2,649 citations262 upvotesRead →Gemini 1.5: Unlocking multimodal understanding across millions of tokens of contextRecalls details across millions of tokens of text, hours of video and days of audio.Google DeepMind4,190 citations65 upvotesRead →Nemotron-4 15B Technical ReportNemotron-4 15B, a large multilingual language model, excels in English, multilingual, and coding tasks, demonstrating superior performance in multilingual capabilities compared to larger and specialized models.NVIDIA36 citations46 upvotesRead →DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code IntelligenceDeepSeek-Coder, a series of open-source code models trained from scratch, outperforms both open-source and closed-source models across multiple benchmarks.
DeepSeek2,033 citations74 upvotesRead →