archive
Every paper Pith has read. Search by title, abstract, or pith.
1286 papers in cs.IR · page 6
-
Citation networks map broad fields for initial reviews
SotA Lens: A Network-Augmented Methodology and Tool for Exploratory State-of-the-Art Reviews
-
Aspect-aware tests expose hidden retriever gaps in reasoning tasks
Rethinking Reasoning-Intensive Retrieval: Evaluating and Advancing Retrievers in Agentic Search Systems
-
Mixed training lifts average scores on Brazilian legal retrieval
Domain-Adaptive Dense Retrieval for Brazilian Legal Search
-
Multi-agent AI lifts tool execution success 87.5 points in machining
Physics-Grounded Multi-Agent Architecture for Traceable, Risk-Aware Human-AI Decision Support in Manufacturing
1 Piths -
Aspect-aware GNN beats baselines on math paper recommendations
Aspect-Aware Content-Based Recommendations for Mathematical Research Papers
-
Decision module cuts over-acceptance in recommender simulators
Decision-aware User Simulation Agent for Evaluating Conversational Recommender Systems
-
Neural retrievers drop below 0.02 recall on controlled queries
Reproducing Complex Set-Compositional Information Retrieval
-
Multimodal alignment embeds visual semantics into semantic IDs
TriAlignGR: Triangular Multitask Alignment with Multimodal Deep Interest Mining for Generative Recommendation
-
Aligning images and interests into SIDs reduces degradation and opacity
TriAlignGR: Triangular Multitask Alignment with Multimodal Deep Interest Mining for Generative Recommendation
-
Python package unifies music performance feature pipelines
Cosmodoit: A Python Package for Adaptive, Efficient Pipelining of Feature Extraction from Performed Music
-
SURE-RAG cuts unsafe answers 37 percent at 30 percent coverage
SURE-RAG: Sufficiency and Uncertainty-Aware Evidence Verification for Selective Retrieval-Augmented Generation
-
One physical sensor can stand in for many with under 3.2% error
Information Density as a Quantitative Measure for AI-enabled Virtual Sensing: Feasibility and Limits
-
GenPOI generates map POIs from context via LLM token sequences
Revisiting General Map Search via Generative Point-of-Interest Retrieval
-
Thinking traces as RAG corpus raise math and code scores
RAG over Thinking Traces Can Improve Reasoning Tasks
-
Unified PostgreSQL layer cuts RAG latency up to 92%
Beyond Similarity Search: A Unified Data Layer for Production RAG Systems
-
VLM infers cues to retrieve album refs for personal photo completion
AlbumFill: Album-Guided Reasoning and Retrieval for Personalized Image Completion
-
Speech embeddings split by attribute for selective similarity searches
Multi-Axis Speech Similarity via Factor-Partitioned Embeddings
-
Partitioned speech vectors allow searches that ignore speaker or focus on words
Multi-Axis Speech Similarity via Factor-Partitioned Embeddings
-
Gap-repair controller raises multi-hop RAG evidence F1 with 2.6x fewer tokens
AdaGATE: Adaptive Gap-Aware Token-Efficient Evidence Assembly for Multi-Hop Retrieval-Augmented Generation
-
Cross-encoder reranking leads biomedical RAG retrieval by small but clear margin
Benchmarking Retrieval Strategies for Biomedical Retrieval-Augmented Generation: A Controlled Empirical Study
-
AI system reconstructs BSM exclusion limits from HEPData
From Experimental Limits to Physical Insight: A Retrieval-Augmented Multi-Agent Framework for Interpreting Searches Beyond the Standard Model
-
GRAIL cuts agent discovery latency by 79x using SLM and resonance
GRAIL: A Deep-Granularity Hybrid Resonance Framework for Real-Time Agent Discovery via SLM-Enhanced Indexing
-
Two-stage filter catches unfaithful RAG answers up to 77 percent of cases
Towards Dependable Retrieval-Augmented Generation Using Factual Confidence Prediction
-
Evolutionary probes improve agent tool retrieval from rank 8.8 to 2.8
FitText: Evolving Agent Tool Ecologies via Memetic Retrieval
-
Feature-level analysis beats claim-level classification for patent novelty
Is It Novel and Why? Fine-Grained Patent Novelty Prediction Based on Passage Retrieval
-
Framework aligns agents for fair multistakeholder personalization
Fair Agents: Balancing Multistakeholder Alignment in Multi-Agent Personalization Systems
-
ODE behavior modeling plus LLM counterfactuals lifts cross-domain recs
Bridging Behavior and Semantics for Time-aware Cross-Domain Sequential Recommendation
-
Direct text tools beat semantic retrievers for agents
Beyond Semantic Similarity: Rethinking Retrieval for Agentic Search via Direct Corpus Interaction
-
Agents and rubric rewards improve AI-generated legal judgments
Enhancing Judgment Document Generation via Agentic Legal Information Collection and Rubric-Guided Optimization
-
Hybrid AI agents on security tools close the alert gap
CyberAId: AI-Driven Cybersecurity for Financial Service Providers
-
Target-aware frequency filtering isolates true user interests from noise
FEDIN: Frequency-Enhanced Deep Interest Network for Click-Through Rate Prediction
-
Hybrid RAG reaches perfect grounding accuracy on 200 biomedical claims
A Hybrid Retrieval and Reranking Framework for Evidence-Grounded Retrieval-Augmented Generation
-
LLM framework tricks neural rankers to promote target documents
Led to Mislead: Adversarial Content Injection for Attacks on Neural Ranking Models
-
Hybrid KG-LLM service retrieves skills at 0.94 nDCG under 200 ms
KG-First, LLM-Fallback: A Hybrid Microservice for Grounded Skill Search and Explanation
-
Dynamic multi-similarity graphs raise recommender recall
Dynamic Graph with Similarity-Aware Attention Graph Neural Network for Recommender Systems
-
Post-hoc adapter steers recommender exposure toward provider fairness
Post-hoc Provider Fairness Adaptation via Hierarchical Exposure Alignment
-
Multi-turn queries sharpen health video retrieval
Interactive Multi-Turn Retrieval for Health Videos
-
General pre-training with higher rates improves SPLADE retrieval
The Pre-Training Study of Expanded-SPLADE Models on Web Document Titles
-
Profile control alone improves perceptions in learning recommenders
Investigating the Effects of Different Levels of User Control in an Interactive Educational Recommender System
-
Verbal reranker adds explicit logic links to RAG
Verbal-R3: Verbal Reranker as the Missing Bridge between Retrieval and Reasoning
-
RAG retrieval prioritizes decision safety over semantic match for biased queries
Beyond Semantic Relevance: Counterfactual Risk Minimization for Robust Retrieval-Augmented Generation
-
Screenshots replace text parsing for traceable evidence in RAG
Chain of Evidence: Pixel-Level Visual Attribution for Iterative Retrieval-Augmented Generation
-
Pair trimming and expert blending shrink modality gap by 90%
Multimodal Data Curation Through Ranked Retrieval
-
Workbench links weather embeddings to physical data for event retrieval
Toward a Scientific Discovery Engine for Weather and Climate Data: A Visual Analytics Workbench for Embedding-Based Exploration
-
Human-AI teams outperform AI alone in info-seeking tasks
Seeking Information with RAG-Assistants: Does Model Size Matter in Human-AI Collaborations?
-
Subgraph retrieval completes missing modalities in recommendations
Robust Multimodal Recommendation via Graph Retrieval-Enhanced Modality Completion
-
XTR training boosts efficiency in IVF retrieval engines
A Replicability Study of XTR
-
Hierarchical RAG reaches 0.427 nDCG on multi-turn retrieval
H-RAG at SemEval-2026 Task 8: Hierarchical Parent-Child Retrieval for Multi-Turn RAG Conversations
-
Relevance scores help RAG systems convey uncertainty
"I Don't Know" -- Towards Appropriate Trust with Certainty-Aware Retrieval Augmented Generation
-
Dual scoring lifts keyphrase extraction accuracy
MUDY: Multi-Granular Dynamic Candidate Contextualization for Unsupervised Keyphrase Extraction