Paperium - Forem

Cover image for Loopholing Discrete Diffusion: Deterministic Bypass of the Sampling Wall

Paperium

Nov 15

Loopholing Discrete Diffusion: Deterministic Bypass of the Sampling Wall

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for HSCodeComp: A Realistic and Expert-level Benchmark for Deep Search Agents inHierarchical Rule Application

Paperium

Nov 15

HSCodeComp: A Realistic and Expert-level Benchmark for Deep Search Agents inHierarchical Rule Application

#ai #deeplearning #computerscience #machinelearning

2 min read

Cover image for DeepWideSearch: Benchmarking Depth and Width in Agentic Information Seeking

Paperium

Nov 14

DeepWideSearch: Benchmarking Depth and Width in Agentic Information Seeking

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Text or Pixels? It Takes Half: On the Token Efficiency of Visual Text Inputs inMultimodal LLMs

Paperium

Nov 14

Text or Pixels? It Takes Half: On the Token Efficiency of Visual Text Inputs inMultimodal LLMs

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Accelerating Vision Transformers with Adaptive Patch Sizes

Paperium

Nov 14

Accelerating Vision Transformers with Adaptive Patch Sizes

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for SAVANT: Semantic Analysis with Vision-Augmented Anomaly deTection

Paperium

Nov 14

SAVANT: Semantic Analysis with Vision-Augmented Anomaly deTection

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Machine Text Detectors are Membership Inference Attacks

Paperium

Nov 14

Machine Text Detectors are Membership Inference Attacks

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for DeLeaker: Dynamic Inference-Time Reweighting For Semantic Leakage Mitigation inText-to-Image Models

Paperium

Nov 14

DeLeaker: Dynamic Inference-Time Reweighting For Semantic Leakage Mitigation inText-to-Image Models

#ai #deeplearning #computerscience #machinelearning

2 min read

Cover image for What Questions Should Robots Be Able to Answer? A Dataset of User Questions forExplainable Robotics

Paperium

Nov 14

What Questions Should Robots Be Able to Answer? A Dataset of User Questions forExplainable Robotics

#ai #deeplearning #computerscience #machinelearning

2 min read

Cover image for RIR-Mega: a large-scale simulated room impulse response dataset for machinelearning and room acoustics modeling

Paperium

Nov 14

RIR-Mega: a large-scale simulated room impulse response dataset for machinelearning and room acoustics modeling

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for See the Text: From Tokenization to Visual Reading

Paperium

Nov 14

See the Text: From Tokenization to Visual Reading

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for When Do Transformers Learn Heuristics for Graph Connectivity?

Paperium

Nov 14

When Do Transformers Learn Heuristics for Graph Connectivity?

#ai #deeplearning #computerscience #machinelearning

2 min read

Cover image for Learning from the Best, Differently: A Diversity-Driven Rethinking on DataSelection

Paperium

Nov 14

Learning from the Best, Differently: A Diversity-Driven Rethinking on DataSelection

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for ProfBench: Multi-Domain Rubrics requiring Professional Knowledge to Answer andJudge

Paperium

Nov 14

ProfBench: Multi-Domain Rubrics requiring Professional Knowledge to Answer andJudge

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Steering Autoregressive Music Generation with Recursive Feature Machines

Paperium

Nov 14

Steering Autoregressive Music Generation with Recursive Feature Machines

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for MINED: Probing and Updating with Multimodal Time-Sensitive Knowledge for LargeMultimodal Models

Paperium

Nov 14

MINED: Probing and Updating with Multimodal Time-Sensitive Knowledge for LargeMultimodal Models

#ai #deeplearning #computerscience #machinelearning

2 min read

Cover image for From Charts to Code: A Hierarchical Benchmark for Multimodal Models

Paperium

Nov 14

From Charts to Code: A Hierarchical Benchmark for Multimodal Models

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for NeuroAda: Activating Each Neuron's Potential for Parameter-Efficient Fine-Tuning

Paperium

Nov 14

NeuroAda: Activating Each Neuron's Potential for Parameter-Efficient Fine-Tuning

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for TheMCPCompany: Creating General-purpose Agents with Task-specific Tools

Paperium

Nov 14

TheMCPCompany: Creating General-purpose Agents with Task-specific Tools

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for ColorAgent: Building A Robust, Personalized, and Interactive OS Agent

Paperium

Nov 14

ColorAgent: Building A Robust, Personalized, and Interactive OS Agent

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for OmniNWM: Omniscient Driving Navigation World Models

Paperium

Nov 14

OmniNWM: Omniscient Driving Navigation World Models

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Are they lovers or friends? Evaluating LLMs' Social Reasoning in English andKorean Dialogues

Paperium

Nov 13

Are they lovers or friends? Evaluating LLMs' Social Reasoning in English andKorean Dialogues

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for KORE: Enhancing Knowledge Injection for Large Multimodal Models viaKnowledge-Oriented Augmentations and Constraints

Paperium

Nov 13

KORE: Enhancing Knowledge Injection for Large Multimodal Models viaKnowledge-Oriented Augmentations and Constraints

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Directional Reasoning Injection for Fine-Tuning MLLMs

Paperium

Nov 13

Directional Reasoning Injection for Fine-Tuning MLLMs

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for FinSight: Towards Real-World Financial Deep Research

Paperium

Nov 13

FinSight: Towards Real-World Financial Deep Research

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Decomposed Attention Fusion in MLLMs for Training-Free Video ReasoningSegmentation

Paperium

Nov 13

Decomposed Attention Fusion in MLLMs for Training-Free Video ReasoningSegmentation

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for olmOCR 2: Unit Test Rewards for Document OCR

Paperium

Nov 13

olmOCR 2: Unit Test Rewards for Document OCR

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Unified Reinforcement and Imitation Learning for Vision-Language Models

Paperium

Nov 13

Unified Reinforcement and Imitation Learning for Vision-Language Models

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Attention Sinks in Diffusion Language Models

Paperium

Nov 13

Attention Sinks in Diffusion Language Models

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Language Models are Injective and Hence Invertible

Paperium

Nov 13

Language Models are Injective and Hence Invertible

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Pico-Banana-400K: A Large-Scale Dataset for Text-Guided Image Editing

Paperium

Nov 13

Pico-Banana-400K: A Large-Scale Dataset for Text-Guided Image Editing

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for VideoAgentTrek: Computer Use Pretraining from Unlabeled Videos

Paperium

Nov 13

VideoAgentTrek: Computer Use Pretraining from Unlabeled Videos

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Towards Faithful and Controllable Personalization via Critique-Post-EditReinforcement Learning

Paperium

Nov 13

Towards Faithful and Controllable Personalization via Critique-Post-EditReinforcement Learning

#ai #deeplearning #computerscience #machinelearning

2 min read

Cover image for Human-Agent Collaborative Paper-to-Page Crafting for Under $0.1

Paperium

Nov 13

Human-Agent Collaborative Paper-to-Page Crafting for Under $0.1

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for ImpossibleBench: Measuring LLMs' Propensity of Exploiting Test Cases

Paperium

Nov 13

ImpossibleBench: Measuring LLMs' Propensity of Exploiting Test Cases

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for GigaBrain-0: A World Model-Powered Vision-Language-Action Model

Paperium

Nov 13

GigaBrain-0: A World Model-Powered Vision-Language-Action Model

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for DaMo: Data Mixing Optimizer in Fine-tuning Multimodal LLMs for Mobile PhoneAgents

Paperium

Nov 13

DaMo: Data Mixing Optimizer in Fine-tuning Multimodal LLMs for Mobile PhoneAgents

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced PolicyOptimization with Adaptive Clipping

Paperium

Nov 13

BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced PolicyOptimization with Adaptive Clipping

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Every Attention Matters: An Efficient Hybrid Architecture for Long-ContextReasoning

Paperium

Nov 13

Every Attention Matters: An Efficient Hybrid Architecture for Long-ContextReasoning

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for LoongRL:Reinforcement Learning for Advanced Reasoning over Long Contexts

Paperium

Nov 13

LoongRL:Reinforcement Learning for Advanced Reasoning over Long Contexts

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for When Correct Is Not Safe: Can We Trust Functionally Correct Patches Generatedby Code Agents?

Paperium

Nov 13

When Correct Is Not Safe: Can We Trust Functionally Correct Patches Generatedby Code Agents?

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Pruning Overparameterized Multi-Task Networks for Degraded Web Image Restoration

Paperium

Nov 13

Pruning Overparameterized Multi-Task Networks for Degraded Web Image Restoration

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for PokeeResearch: Effective Deep Research via Reinforcement Learning from AIFeedback and Robust Reasoning Scaffold

Paperium

Nov 12

PokeeResearch: Effective Deep Research via Reinforcement Learning from AIFeedback and Robust Reasoning Scaffold

#ai #deeplearning #computerscience #machinelearning

2 min read

Cover image for Static Sandboxes Are Inadequate: Modeling Societal Complexity RequiresOpen-Ended Co-Evolution in LLM-Based Multi-Agent Simulatio

Paperium

Nov 12

Static Sandboxes Are Inadequate: Modeling Societal Complexity RequiresOpen-Ended Co-Evolution in LLM-Based Multi-Agent Simulatio

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Predicting the Unpredictable: Reproducible BiLSTM Forecasting of Incident Countsin the Global Terrorism Database (GTD)

Paperium

Nov 12

Predicting the Unpredictable: Reproducible BiLSTM Forecasting of Incident Countsin the Global Terrorism Database (GTD)

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Unimedvl: Unifying Medical Multimodal Understanding And Generation ThroughObservation-Knowledge-Analysis

Paperium

Nov 12

Unimedvl: Unifying Medical Multimodal Understanding And Generation ThroughObservation-Knowledge-Analysis

#ai #deeplearning #computerscience #machinelearning

1 min read

Paperium

Nov 12

Planned Diffusion

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Expanding the Action Space of LLMs to Reason Beyond Language

Paperium

Nov 12

Expanding the Action Space of LLMs to Reason Beyond Language

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Any-Depth Alignment: Unlocking Innate Safety Alignment of LLMs to Any-Depth

Paperium

Nov 12

Any-Depth Alignment: Unlocking Innate Safety Alignment of LLMs to Any-Depth

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Think with 3D: Geometric Imagination Grounded Spatial Reasoning from LimitedViews

Paperium

Nov 12

Think with 3D: Geometric Imagination Grounded Spatial Reasoning from LimitedViews

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for DeepSeek-OCR: Contexts Optical Compression

Paperium

Nov 12

DeepSeek-OCR: Contexts Optical Compression

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Is Multilingual LLM Watermarking Truly Multilingual? A Simple Back-TranslationSolution

Paperium

Nov 12

Is Multilingual LLM Watermarking Truly Multilingual? A Simple Back-TranslationSolution

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for GAS: Improving Discretization of Diffusion ODEs via Generalized AdversarialSolver

Paperium

Nov 12

GAS: Improving Discretization of Diffusion ODEs via Generalized AdversarialSolver

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Efficient Long-context Language Model Training by Core Attention Disaggregation

Paperium

Nov 12

Efficient Long-context Language Model Training by Core Attention Disaggregation

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for EvoSyn: Generalizable Evolutionary Data Synthesis for Verifiable Learning

Paperium

Nov 12

EvoSyn: Generalizable Evolutionary Data Synthesis for Verifiable Learning

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Extracting alignment data in open models

Paperium

Nov 12

Extracting alignment data in open models

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for AlphaQuanter: An End-to-End Tool-Orchestrated Agentic Reinforcement LearningFramework for Stock Trading

Paperium

Nov 12

AlphaQuanter: An End-to-End Tool-Orchestrated Agentic Reinforcement LearningFramework for Stock Trading

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for PRISMM-Bench: A Benchmark of Peer-Review Grounded Multimodal Inconsistencies

Paperium

Nov 12

PRISMM-Bench: A Benchmark of Peer-Review Grounded Multimodal Inconsistencies

#ai #deeplearning #computerscience #machinelearning

1 min read

Cover image for Mono4DGS-HDR: High Dynamic Range 4D Gaussian Splatting from Alternating-exposureMonocular Videos

Paperium

Nov 12

Mono4DGS-HDR: High Dynamic Range 4D Gaussian Splatting from Alternating-exposureMonocular Videos

#ai #deeplearning #computerscience #machinelearning

2 min read

Cover image for Video Reasoning without Training

Paperium

Nov 12

Video Reasoning without Training

#ai #deeplearning #computerscience #machinelearning

1 min read