Recent Research Signals

Found 431 results

Filtering papers...

Paper	Possible Impact	Features	Possible Business Area
Aggregating maximal cliques in real-world graphs Data Structures and Algorithms	Finds important groups in data faster.	Google Princeton University Microsoft 2 Countries	Big Data (73%)	Score: 5 2 days ago
Privacy-Preserving Generative Modeling and Clinical Validation of Longitudinal Health Records for Chronic Disease Machine Learning (CS)	Creates fake patient data to train medical AI.	Stanford University Code 2 Countries	Predictive Analytics (78%)	Score: 4 6 days ago
The Trojan Knowledge: Bypassing Commercial LLM Guardrails via Harmless Prompt Weaving and Adaptive Tree Search Cryptography and Security	Breaks AI safety by tricking it with hidden questions.	IBM Code 2 Countries	Natural Language Processing (76%)	Score: 4 4 days ago
InvertiTune: High-Quality Data Synthesis for Cost-Effective Single-Shot Text-to-Knowledge Graph Generation Computation and Language	Teaches computers to build knowledge maps from text.	Huawei Data 2 Countries	Text Analytics (80%)	Score: 4 3 days ago
MoReGen: Multi-Agent Motion-Reasoning Engine for Code-based Text-to-Video Synthesis CV and Pattern Recognition	Makes videos follow real-world physics rules.	Microsoft Code 2 Countries	Motion Capture (80%)	Score: 4 2 days ago
Distill, Forget, Repeat: A Framework for Continual Unlearning in Text-to-Image Diffusion Models Machine Learning (CS)	Removes unwanted data from AI without retraining.	Sony PlayStation Code 2 Countries	Machine Learning (77%)	Score: 4 3 days ago
CycliST: A Video Language Model Benchmark for Reasoning on Cyclical State Transitions CV and Pattern Recognition	Teaches computers to understand repeating patterns in videos.	Code Data 2 Countries	Image Recognition (77%)	Score: 4 5 days ago
LLM CHESS: Benchmarking Reasoning and Instruction-Following in LLMs through Chess Artificial Intelligence	Tests how well AI plays and understands chess.	University of California, Berkeley Code	Machine Learning (78%)	Score: 3 4 days ago
A Wolf in Sheep's Clothing: Bypassing Commercial LLM Guardrails via Harmless Prompt Weaving and Adaptive Tree Search Cryptography and Security	Breaks AI safety rules by tricking its knowledge.	Code 2 Countries	Natural Language Processing (76%)	Score: 3 4 days ago
Rectifying LLM Thought from Lens of Optimization Computation and Language	Teaches computers to think smarter, not longer.	Code Data	Natural Language Processing (78%)	Score: 3 4 days ago
CAR-Net: A Cascade Refinement Network for Rotational Motion Deblurring under Angle Information Uncertainty CV and Pattern Recognition	Fixes blurry pictures from spinning.	Code 2 Countries	Image Recognition (76%)	Score: 3 6 days ago
Towards Irreversible Machine Unlearning for Diffusion Models Machine Learning (CS)	Makes AI forget unwanted pictures, safely.	Code 2 Countries	Machine Learning (76%)	Score: 3 2 days ago
CogDrive: Cognition-Driven Multimodal Prediction-Planning Fusion for Safe Autonomy Robotics	Helps self-driving cars predict and avoid accidents.	Massachusetts Institute of Technology 3 Countries	Autonomous Vehicles (81%)	Score: 3 3 days ago
Four Over Six: More Accurate NVFP4 Quantization with Adaptive Block Scaling Computation and Language	Makes AI models train faster and better.	Massachusetts Institute of Technology Code	Quantum Computing (73%)	Score: 3 4 days ago
PhyDetEx: Detecting and Explaining the Physical Plausibility of T2V Models CV and Pattern Recognition	Teaches AI to spot videos breaking physics rules.	Code 2 Countries	Image Recognition (77%)	Score: 3 4 days ago
UserSimCRS v2: Simulation-Based Evaluation for Conversational Recommender Systems Information Retrieval	Builds better chatbots that suggest things you like.	Code 2 Countries	Simulation (80%)	Score: 3 1 day ago
Equivariant Symmetry-Aware Head Pose Estimation for Fetal MRI CV and Pattern Recognition	Helps doctors see babies better during scans.	Massachusetts Institute of Technology Code	Image Recognition (75%)	Score: 3 1 day ago
Lumos: Let there be Language Model System Certification Programming Languages	Finds dangerous mistakes in AI that drives cars.	Code 2 Countries	Natural Language Processing (77%)	Score: 3 3 days ago
AdaptVision: Efficient Vision-Language Models via Adaptive Visual Acquisition CV and Pattern Recognition	Lets computers see smarter, using less data.	Code Data	Image Recognition (80%)	Score: 3 2 days ago
CoDA: From Text-to-Image Diffusion Models to Training-Free Dataset Distillation CV and Pattern Recognition	Makes AI learn from less data, faster.	Code 2 Countries	Image Recognition (76%)	Score: 3 2 days ago
SEASON: Mitigating Temporal Hallucination in Video Large Language Models via Self-Diagnostic Contrastive Decoding CV and Pattern Recognition	Fixes videos so AI understands time better.	NVIDIA 2 Countries	Visual Search (75%)	Score: 3 1 day ago
Memory-Integrated Reconfigurable Adapters: A Unified Framework for Settings with Multiple Tasks Machine Learning (CS)	AI learns new things without forgetting old ones.	Microsoft 2 Countries	Intelligent Systems (78%)	Score: 3 5 days ago
TRoVe: Discovering Error-Inducing Static Feature Biases in Temporal Vision-Language Models CV and Pattern Recognition	Finds hidden mistakes in AI that watches videos.	Stanford University Code	Image Recognition (77%)	Score: 3 5 days ago
UniCom: Towards a Unified and Cohesiveness-aware Framework for Community Search and Detection Social and Information Networks	Finds groups in networks without needing new training.	Code 2 Countries	Semantic Search (79%)	Score: 3 3 days ago
Spatiotemporal Pyramid Flow Matching for Climate Emulation CV and Pattern Recognition	Models climate change faster and more accurately.	Stanford University Code	Simulation (77%)	Score: 3 4 days ago
Four Over Six: More Accurate NVFP4 Quantization with Adaptive Block Scaling Computation and Language	Trains AI faster and better using less computer memory.	Massachusetts Institute of Technology Code	A/B Testing (73%)	Score: 3 4 days ago
Layered Division and Global Allocation for Community Detection in Multilayer Network Social and Information Networks	Finds hidden groups in connected data.	Massachusetts Institute of Technology 2 Countries	Content Delivery Network (77%)	Score: 3 4 days ago
GalaxyDiT: Efficient Video Generation with Guidance Alignment and Adaptive Proxy in Diffusion Transformers CV and Pattern Recognition	Makes videos faster without losing quality.	NVIDIA Massachusetts Institute of Technology	Image Recognition (76%)	Score: 3 2 days ago
Generative AI for Self-Adaptive Systems: State of the Art and Research Roadmap Software Engineering	Makes smart systems learn and adapt better.	Code 3 Countries	Intelligent Systems (83%)	Score: 3 1 day ago
TUNA: Taming Unified Visual Representations for Native Unified Multimodal Models CV and Pattern Recognition	Lets computers understand and create pictures and videos.	Meta 3 Countries	Unified Communications (76%)	Score: 3 4 days ago
SAM3-UNet: Simplified Adaptation of Segment Anything Model 3 CV and Pattern Recognition	Teaches computers to find things in pictures faster.	Code	Image Recognition (76%)	Score: 2 4 days ago
Bias Testing and Mitigation in Black Box LLMs using Metamorphic Relations Software Engineering	Finds and fixes hidden unfairness in AI.	2 Countries	A/B Testing (77%)	Score: 2 6 days ago
OpenREAD: Reinforced Open-Ended Reasoning for End-to-End Autonomous Driving with LLM-as-Critic CV and Pattern Recognition	Helps self-driving cars think and plan better.	Code	Autonomous Vehicles (79%)	Score: 2 4 days ago
Probabilistic Modeling of Multi-rater Medical Image Segmentation for Diversity and Personalization CV and Pattern Recognition	Makes medical scans show different expert opinions.	Code	Image Recognition (78%)	Score: 2 5 days ago
IndicParam: Benchmark to evaluate LLMs on low-resource Indic Languages Computation and Language	Helps computers understand rare Indian languages better.	Code	Natural Language Processing (76%)	Score: 2 6 days ago
MSPT: Efficient Large-Scale Physical Modeling via Parallelized Multi-Scale Attention Machine Learning (CS)	Solves hard science problems faster on computers.	Code	Natural Language Processing (75%)	Score: 2 4 days ago
OpenREAD: Reinforced Open-Ended Reasoing for End-to-End Autonomous Driving with LLM-as-Critic CV and Pattern Recognition	Teaches cars to think and drive better.	Code	Autonomous Vehicles (80%)	Score: 2 4 days ago
Deep FlexQP: Accelerated Nonlinear Programming via Deep Unfolding Optimization and Control	Solves hard math problems faster, even if they seem impossible.	Massachusetts Institute of Technology	Quantum Computing (75%)	Score: 2 4 days ago
CourtMotion: Learning Event-Driven Motion Representations from Skeletal Data for Basketball CV and Pattern Recognition	Predicts basketball plays by watching player movements.	Amazon	Motion Capture (83%)	Score: 2 4 days ago
Scaling and context steer LLMs along the same computational path as the human brain Machine Learning (CS)	Brain and AI process information in a similar order.	Code	Natural Language Processing (79%)	Score: 2 4 days ago
Flow Matching for Tabular Data Synthesis Machine Learning (CS)	Creates private fake data that's better than old ways.	2 Countries	A/B Testing (75%)	Score: 2 6 days ago
MEGConformer: Conformer-Based MEG Decoder for Robust Speech and Phoneme Classification Computation and Language	Helps computers understand spoken words from brain waves.	Code	Speech Recognition (75%)	Score: 2 4 days ago
Stay Unique, Stay Efficient: Preserving Model Personality in Multi-Task Merging Machine Learning (CS)	Combines AI skills without losing what it learned.	Code	Personalization (77%)	Score: 2 4 days ago
GPTrace: Effective Crash Deduplication Using LLM Embeddings Software Engineering	Finds duplicate software bugs faster.	Code	Text Analytics (75%)	Score: 2 4 days ago
Tangram: Accelerating Serverless LLM Loading through GPU Memory Reuse and Affinity Distributed, Parallel, and Cluster Computing	Makes AI models load much faster for users.	2 Countries	Cloud Computing (76%)	Score: 2 4 days ago
Multilingual Training-Free Remote Sensing Image Captioning CV and Pattern Recognition	Lets computers describe satellite pictures in any language.	Code	Image Recognition (80%)	Score: 2 5 days ago
Fourier Neural Operators Explained: A Practical Perspective Machine Learning (CS)	Teaches computers to solve hard math problems faster.	Code	Natural Language Processing (76%)	Score: 2 4 days ago
RECTor: Robust and Efficient Correlation Attack on Tor Cryptography and Security	Makes Tor harder to spy on online.	Code	Darknet (75%)	Score: 2 6 days ago
Look, Recite, Then Answer: Enhancing VLM Performance via Self-Generated Knowledge Hints CV and Pattern Recognition	Helps computers see plants better, not guess.	Code	Visual Search (79%)	Score: 2 5 days ago
MDiff4STR: Mask Diffusion Model for Scene Text Recognition CV and Pattern Recognition	Reads messy text faster and more accurately.	Code	Image Recognition (74%)	Score: 2 4 days ago