archive
Every paper Pith has read. Search by title, abstract, or pith.
14513 papers in cs.AI · page 20
-
Varying environment rules builds agents that generalize
Scalable Environments Drive Generalizable Agents
-
Agentic selector ranks second on four-day multimodal challenge
MARS: Technical Report for the CASTLE Challenge at EgoVis 2026
-
Proxy images from EEG let AI models interpret brain signals
Visualizing the Invisible: Generative Visual Grounding Empowers Universal EEG Understanding in MLLMs
-
One universal fix reduces hallucinations in 15 models
TRACE: Trajectory Correction from Cross-layer Evidence for Hallucination Reduction
-
Consistency reward lifts VLM spatial reasoning
Self-Evolving Spatial Reasoning in Vision Language Models via Geometric Logic Consistency
-
New module keeps multimodal models true to images during long answers
Vision Inference Former: Sustaining Visual Consistency in Multimodal Large Language Models
-
Black-box agents revive erased concepts in image generators
Whispers in the Noise: Surrogate-Guided Concept Awakening via a Multi-Agent Framework
-
pArticleMap recovers 10.8% of future papers from literature gaps
Evidence-Grounded Frontier Mapping and Agentic Hypothesis Generation in Nanomedicine
-
GenAI lifts average performance but splits gains by interaction skill
Generative AI and the Productivity Divide: Human-AI Complementarities in Education
-
Indirect injections hijack chatbots to leak user data
An Empirical Study of Privacy Leakage Chains via Prompt Injection in Black-Box Chatbot Environments
-
3D generators leave fingerprints that identify their source
Who Generated This 3D Asset? Learning Source Attribution for Generative 3D Models
-
Adversarial priors raise recall in multivariate anomaly detection
POST: Prior-Observation Adversarial Learning of Spatio-Temporal Associations for Multivariate Time Series Anomaly Detection
-
Grounding cuts tokens 18x while matching big models on home tasks
TaskGround: Structured Executable Task Inference for Full-Scene Household Reasoning
-
Symmetry-respecting updates beat AdamW in LLM pretraining
Symmetry-Compatible Principle for Optimizer Design: Embeddings, LM Heads, SwiGLU MLPs, and MoE Routers
-
Modality drift collapses refusal geometry in multimodal LLMs
Safety Geometry Collapse in Multimodal LLMs and Adaptive Drift Correction
-
Diffusion model generates aligned urban energy maps from roads
SENSE: Satellite-based ENergy Synthesis for Sustainable Environment
-
Attention plus contrastive learning solves mixed-geometry routing
Learning to Solve Compositional Geometry Routing Problems
-
SynGR boosts generative recs by limiting dominant modalities
SynGR: Unleashing the Potential of Cross-Modal Synergy for Generative Recommendation
-
4-qubit game circuit forecasts disruptive capital trajectories
Parameterized 4-Qubit EWL Quantum Game Circuits with Dirac-Solow-Swan Hamiltonian Integration for Quadruple Helix Disruptive Innovation Recommender Systems
-
LLM refines protocols so agents reconstruct states uniformly
LLM-Guided Communication for Cooperative Multi-Agent Reinforcement Learning
-
Multi-model feedback doubles AI solves on contest problems
A-ProS: Towards Reliable Autonomous Programming Through Multi-Model Feedback
-
Curvature rewiring cuts over-squashing in forecasting models
Improving Spatio-Temporal Residual Error Propagation by Mitigating Over-Squashing
-
New model predicts gene expression from tissue slides with better structure
FLAG: Foundation model representation with Latent diffusion Alignment via Graph for spatial gene expression prediction
-
GUI agents search docs for rare tasks
DocOS: Towards Proactive Document-Guided Actions in GUI Agents
-
Softmax uncertainty performs like ensembles for robot gating decisions
Confidence-Gated Robot Autonomy: When Does Uncertainty Actually Help?
-
ProcBench detects process defects in LLM coding agents missed by outcome scores
ProcCtrlBench: Evaluating Process-Level Defects and Control Preservation in LLM Coding Agents
-
Process benchmark catches mid-task defects in LLM coding agents
ProcCtrlBench: Evaluating Process-Level Defects and Control Preservation in LLM Coding Agents
-
Limitation disclosures calibrate case-by-case trust in XAI
Exploring Trust Calibration in XAI - The Impact of Exposing Model Limitations to Lay Users
-
Variance reduction lifts ZO hard-thresholding direction limit
New Insight of Variance reduce in Zero-Order Hard-Thresholding: Mitigating Gradient Error and Expansivity Contradictions
-
Tool localizes node errors in multi-agent LLM workflows
PROTEA: Offline Evaluation and Iterative Refinement for Multi-Agent LLM Workflows
-
Quantum sidecars generate signals for AI optimizers
Quantum Sidecar Architectures for Hybrid AI Training and Inference: Stateful Protected Registers, Stateless Reset-and-Reprepare Circuits and Quantum Weight-State Outlook
-
Rectification LoRA fixes hallucinations in federated self-distillation
FedSDR: Federated Self-Distillation with Rectification
-
LLMs reach 90% on telecom language but only 30% on fixes
TeleCom-Bench: How Far Are Large Language Models from Industrial Telecommunication Applications?
-
Framework trains agents to coordinate despite disrupted interactions
Interaction-Breaking Adversarial Learning Framework for Robust Multi-Agent Reinforcement Learning
-
Frequency extraction recovers hidden generalization at 80% noise
Unveiling Memorization-Generalization Coexistence: A Case Study on Arithmetic Tasks with Label Noise
-
Training fixes attention so text alone locates video objects
See What I Mean: Aligning Vision and Language Representations for Video Fine-grained Object Understanding
-
TinySAM 2 cuts SAM 2 memory tokens to 7 percent at 90 percent accuracy
TinySAM 2: Extreme Memory Compression for Efficient Track Anything Model
-
Semantic scoring refines distilled image datasets
SAS: Semantic-aware Sampling for Generative Dataset Distillation
-
FPGA accelerator adds on-device learning to spiking networks
Spiker-LL: An Energy-Efficient FPGA Accelerator Enabling Adaptive Local Learning in Spiking Neural Networks
-
Shared backbone PPO outperforms standard in multi-UAV coverage
Shared Backbone PPO for Multi-UAV Communication Coverage with Connection Preservation
-
Verify gate renders multi-agent completions inspectable and fail-closed
Verify-Gated Completion as Admission Control in a Governed Multi-Agent Runtime: A Bounded Architecture Case Study
-
Verify gate turns agent completion into inspectable admission control
Verify-Gated Completion as Admission Control in a Governed Multi-Agent Runtime: A Bounded Architecture Case Study
-
Per-module scaling lifts low-bit quantization accuracy
MARR: Module-Adaptive Residual Reconstruction for Low-Bit Post-Training Quantization
-
E-commerce search lifts new-item GMV 5.3 percent via long-term value estimates
Towards Sustainable Growth: A Multi-Value-Aware Retrieval Framework for E-Commerce Search
-
-
Predictive prefetching cuts RAG latency up to 43.5%
Predictive Prefetching for Retrieval-Augmented Generation
-
New benchmark finds 11-30% indirect prompt injection success in AI agents
LivePI: More Realistic Benchmarking of Agents Against Indirect Prompt Injection
-
Sensitivity-aware SVD compresses physics models at high ratios
SAFE-SVD: Sensitivity-Aware Fidelity-Enforcing SVD for Physics Foundation Models
-
LLM search discovers top kernels for high-dimensional BO
Automated Kernel Discovery Towards Understanding High-dimensional Bayesian Optimization
-
LLMs guide Bayesian optimization to 90% performance in 6 iterations
Unleashing LLMs in Bayesian Optimization: Preference-Guided Framework for Scientific Discovery