Search

"generalization"20 resultsKeyword

V-Retrver: Evidence-Driven Agentic Reasoning for Universal Multimodal Retrieval

Dongyang Chen, Chaoyang Wang et al.Feb 5arXiv

V-Retrver is a new way for AI to search across text and images by double-checking tiny visual details instead of only guessing from words.

#V-Retrver#multimodal retrieval#agentic reasoning

ProAct: Agentic Lookahead in Interactive Environments

Intermediate

Yangbin Yu, Mingyu Yang et al.Feb 5arXiv

ProAct teaches AI agents to think ahead accurately without needing expensive search every time they act.

#ProAct#GLAD#MC-Critic

CoBA-RL: Capability-Oriented Budget Allocation for Reinforcement Learning in LLMs

Intermediate

Zhiyuan Yao, Yi-Kai Zhang et al.Feb 3arXiv

Large language models learn better when we spend more practice time on the right questions at the right moments.

#Reinforcement Learning#RLVR#GRPO

Balancing Understanding and Generation in Discrete Diffusion Models

Intermediate

Yue Liu, Yuzhong Zhao et al.Feb 1arXiv

This paper introduces XDLM, a single model that blends two popular diffusion styles (masked and uniform) so it both understands and generates text and images well.

#XDLM#discrete diffusion#stationary noise kernel

Toward Cognitive Supersensing in Multimodal Large Language Model

Intermediate

Boyi Li, Yifan Shen et al.Feb 2arXiv

This paper teaches multimodal AI models to not just read pictures but to also imagine and think with pictures inside their heads.

#multimodal large language model#visual cognition#latent visual imagery

LLM-in-Sandbox Elicits General Agentic Intelligence

Beginner

Daixuan Cheng, Shaohan Huang et al.Jan 22arXiv

This paper shows that giving an AI a safe, tiny virtual computer (a sandbox) lets it solve many kinds of problems better, not just coding ones.

#LLM-in-Sandbox#Agentic Intelligence#Reinforcement Learning

LongCat-Flash-Thinking-2601 Technical Report

Beginner

Meituan LongCat Team, Anchun Gui et al.Jan 23arXiv

LongCat-Flash-Thinking-2601 is a huge 560-billion-parameter Mixture-of-Experts model built to act like a careful helper that can use tools, browse, code, and solve multi-step tasks.

#Agentic reasoning#Mixture-of-Experts#Asynchronous reinforcement learning

Can LLMs Clean Up Your Mess? A Survey of Application-Ready Data Preparation with LLMs

Intermediate

Wei Zhou, Jun Zhou et al.Jan 22arXiv

This survey explains how large language models (LLMs) can clean, connect, and enrich messy data so it’s ready for real apps like dashboards, fraud detection, and training AI.

#Data Preparation#Data Cleaning#Data Integration

Fast KVzip: Efficient and Accurate LLM Inference with Gated KV Eviction

Intermediate

Jang-Hyun Kim, Dongyoon Han et al.Jan 25arXiv

Fast KVzip is a new way to shrink an LLM’s memory (the KV cache) while keeping answers just as accurate.

#KV cache compression#gated KV eviction#sink attention

AdaReasoner: Dynamic Tool Orchestration for Iterative Visual Reasoning

Intermediate

Mingyang Song, Haoyu Sun et al.Jan 26arXiv

AdaReasoner teaches AI to pick the right visual tools, use them in the right order, and stop using them when they aren’t helping.

#AdaReasoner#dynamic tool orchestration#multimodal large language models

Spark: Strategic Policy-Aware Exploration via Dynamic Branching for Long-Horizon Agentic Learning

Intermediate

Jinyang Wu, Shuo Yang et al.Jan 28arXiv

SPARK is a new way to train AI agents that saves compute by exploring more only at the most important moments.

#SPARK#dynamic branching#strategic exploration

Llama-3.1-FoundationAI-SecurityLLM-Reasoning-8B Technical Report

Beginner

Zhuoran Yang, Ed Li et al.Jan 28arXiv

This paper introduces Foundation-Sec-8B-Reasoning, a small (8 billion parameter) AI model that is trained to “think out loud” before answering cybersecurity questions.

#native reasoning#cybersecurity LLM#chain-of-thought

Hybrid Linear Attention Done Right: Efficient Distillation and Effective Architectures for Extremely Long Contexts

Intermediate

Yingfa Chen, Zhen Leng Thai et al.Jan 29arXiv

This paper shows how to turn a big Transformer model into a faster hybrid model that mixes attention and RNN layers using far less training data (about 2.3B tokens).

#hybrid attention#RNN attention hybrid#linear attention

FutureOmni: Evaluating Future Forecasting from Omni-Modal Context for Multimodal LLMs

Intermediate

Qian Chen, Jinlan Fu et al.Jan 20arXiv

FutureOmni is the first benchmark that tests if multimodal AI models can predict what happens next from both sound and video, not just explain what already happened.

#multimodal LLM#audio-visual reasoning#future forecasting

Atlas: Orchestrating Heterogeneous Models and Tools for Multi-Domain Complex Reasoning

Beginner

Jinyang Wu, Guocheng Zhai et al.Jan 7arXiv

ATLAS is a system that picks the best mix of AI models and helper tools for each question, instead of using just one model or a fixed tool plan.

#ATLAS#LLM routing#tool augmentation

Learnable Multipliers: Freeing the Scale of Language Model Matrix Layers

Intermediate

Maksim Velikanov, Ilyas Chahed et al.Jan 8arXiv

The paper shows that big language models often get stuck with weight sizes set by training hyperparameters instead of by the data, which quietly hurts performance.

#learnable multipliers#weight decay#noise–WD equilibrium

Unlocking Implicit Experience: Synthesizing Tool-Use Trajectories from Text

Intermediate

Zhihao Xu, Rumei Li et al.Jan 15arXiv

The paper shows a new way to teach AI assistants how to use tools in many-step conversations by mining ordinary text on the internet for step-by-step “how-to” knowledge.

#GEM pipeline#text-based trajectory generation#tool-use data synthesis

Aligning Text, Code, and Vision: A Multi-Objective Reinforcement Learning Framework for Text-to-Visualization

Intermediate

Mizanur Rahman, Mohammed Saidul Islam et al.Jan 8arXiv

This paper teaches a model to turn a question about a table into both a short answer and a clear, correct chart.

#Text-to-Visualization#Reinforcement Learning#GRPO

JudgeRLVR: Judge First, Generate Second for Efficient Reasoning

Intermediate

Jiangshan Duo, Hanyu Li et al.Jan 13arXiv

JudgeRLVR teaches a model to be a strict judge of answers before it learns to generate them, which trims bad ideas early.

#RLVR#judge-then-generate#discriminative supervision

Meta-RL Induces Exploration in Language Agents

Intermediate

Yulun Jiang, Liangze Jiang et al.Dec 18arXiv

This paper introduces LAMER, a Meta-RL training framework that teaches language agents to explore first and then use what they learned to solve tasks faster.

#Meta-Reinforcement Learning#Language Agents#Exploration vs Exploitation