A benchmark and ontology-driven framework links 434 cardiovascular devices to patents at 91.6% recall, producing 6.8M high-confidence links for regulatory-IP integration.
hub Canonical reference
Passage Re-ranking with BERT
Canonical reference. 88% of citing Pith papers cite this work as background.
92
Pith papers citing it
21
external citations · Pith
Background
88% of classified citations
abstract
Recently, neural models pretrained on a language modeling task, such as ELMo (Peters et al., 2017), OpenAI GPT (Radford et al., 2018), and BERT (Devlin et al., 2018), have achieved impressive results on various natural language processing tasks such as question-answering and natural language inference. In this paper, we describe a simple re-implementation of BERT for query-based passage re-ranking. Our system is the state of the art on the TREC-CAR dataset and the top entry in the leaderboard of the MS MARCO passage retrieval task, outperforming the previous state of the art by 27% (relative) in MRR@10. The code to reproduce our results is available at https://github.com/nyu-dl/dl4marco-bert
hub tools
citation-role summary
background 15
baseline 1
citation-polarity summary
claims ledger
- abstract Recently, neural models pretrained on a language modeling task, such as ELMo (Peters et al., 2017), OpenAI GPT (Radford et al., 2018), and BERT (Devlin et al., 2018), have achieved impressive results on various natural language processing tasks such as question-answering and natural language inference. In this paper, we describe a simple re-implementation of BERT for query-based passage re-ranking. Our system is the state of the art on the TREC-CAR dataset and the top entry in the leaderboard of the MS MARCO passage retrieval task, outperforming the previous state of the art by 27% (relative)
co-cited works
representative citing papers
The first SoK on LLM-based AutoPT frameworks provides a six-dimension taxonomy of agent designs and a unified empirical benchmark evaluating 15 frameworks via over 10 billion tokens and 1,500 manually reviewed logs.
A self-supervised transformer learns to unscramble Feynman integrals for online IBP reduction, delivering bounded memory use on complex two-loop topologies while matching Kira's speed on the hardest cases tested.
BEIR is a heterogeneous zero-shot benchmark showing BM25 as a robust baseline while re-ranking and late-interaction models perform best on average at higher cost, with dense and sparse models lagging in generalization.
Dense dual-encoder retrievers outperform BM25 by 9-19% absolute in top-20 passage retrieval accuracy across open-domain QA datasets and enable new state-of-the-art end-to-end QA results.
HAKARI-Bench reconstructs 35 benchmarks into 551 tasks across 43 languages, reproducing full MTEB, MMTEB, and BEIR rankings with Spearman correlation above 0.97 while supporting efficiency variant comparisons.
Budgeted selection from temporal evidence graphs raises leakage-safe prospective recall@50 for CVE exploitability from 0.010 to 0.026 on 12012 CVEs, with two documents capturing most value and standard splits inflating performance by 8.5x.
NNN decoding selects documents via non-negative elastic net reconstruction of the query embedding, with a theorem showing it strictly dominates dense retrieval on correlated corpora and experiments showing gains over inner-product baselines.
MetaSyn is a stage-level benchmark of 442 meta-analyses showing LLM agents retrieve up to 90.9% of eligible studies but include at most 52.7% in their final reports.
An adaptive two-phase semantic filter using clustering then a hybrid proxy trained on LLM confidence achieves 1.6-2.0x speedup over prior methods at 90% accuracy on 10K document corpora.
Re-ranking retrieval candidates via a cross-encoder trained on continuous perturbation-based attribution scores improves citation faithfulness and gold-answer alignment in legal QA over semantic similarity.
DART adapts a scoring matrix at inference time via gradient updates on pseudo-labels from top/bottom documents to gain +2.1% mean NDCG@10 on six BEIR benchmarks with under 10ms added latency.
SilentRetrieval is a data poisoning attack achieving 84.6% HR@10 and 57.5% ASR-LLM on Natural Questions via coordinated beam search and trigger fusion while preserving document fluency.
Layer-wise Token Compression applies adaptive token pooling at middle transformer layers for cross-encoder rerankers, preserving MS MARCO ranking quality while raising QPS up to 25% on passages and 116% on documents, with added gains on listwise LLM rerankers and a regularizer effect for long inputs
ZipRerank delivers state-of-the-art multimodal listwise reranking accuracy for long documents at up to 10x lower latency via early interaction and single-pass scoring.
HDRI is a six-principle eight-stage framework for hypothesis-organized LLM research featuring gap-driven iteration, traceable fact reasoning, and subject locking, realized in INFOMINER with reported gains in fact density and completeness.
Prism-Reranker models output relevance, contribution statements, and evidence passages to support agentic retrieval beyond scalar scoring.
BAGEL is a Bayesian active learning framework that uses Gaussian Processes to propagate LLM relevance signals across embedding space and guide global exploration, outperforming standard LLM reranking under identical budgets on four retrieval benchmarks.
KIRA is a unified architecture for visual RAG that reports 0.97 retrieval precision, 1.0 grounding, and 0.707 domain correctness across medical, circuit, satellite, and histopathology domains via hierarchical chunking, dual-path retrieval, and evidence-conditioned generation.
Cross-encoder reranker performance scales predictably via power laws with model size and training exposure, allowing accurate forecasts for 400M and 1B models and data-heavy compute allocation.
SPIRE presents a tree-structured retrieval method using subdocuments, paths, and dual contextualization that produces higher-quality and more diverse citations than passage-based baselines on HTML QA benchmarks.
RAG models set new state-of-the-art results on open-domain QA by retrieving Wikipedia passages and conditioning a generative model on them, while also producing more factual text than parametric baselines.
Structured retrieval (NAVEMBED/NAVINDEX) matches full-corpus injection on legal multi-document QA while cutting tokens 1.6–30×, with a closed-form cache-crossover cost rule.
Governed selection over versioned, hash-chained knowledge vaults Pareto-dominates BM25 on a stale-version attack and is fully deterministic where dense+HNSW retrieval is not.
citing papers explorer
-
From Regulatory Approvals to Patents: Cross-Domain Linking for Cardiovascular Device Traceability
A benchmark and ontology-driven framework links 434 cardiovascular devices to patents at 91.6% recall, producing 6.8M high-confidence links for regulatory-IP integration.
-
Hackers or Hallucinators? A Comprehensive Analysis of LLM-Based Automated Penetration Testing
The first SoK on LLM-based AutoPT frameworks provides a six-dimension taxonomy of agent designs and a unified empirical benchmark evaluating 15 frameworks via over 10 billion tokens and 1,500 manually reviewed logs.
-
Learning to Unscramble Feynman Loop Integrals with SAILIR
A self-supervised transformer learns to unscramble Feynman integrals for online IBP reduction, delivering bounded memory use on complex two-loop topologies while matching Kira's speed on the hardest cases tested.
-
BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Models
BEIR is a heterogeneous zero-shot benchmark showing BM25 as a robust baseline while re-ranking and late-interaction models perform best on average at higher cost, with dense and sparse models lagging in generalization.
-
Dense Passage Retrieval for Open-Domain Question Answering
Dense dual-encoder retrievers outperform BM25 by 9-19% absolute in top-20 passage retrieval accuracy across open-domain QA datasets and enable new state-of-the-art end-to-end QA results.
-
HAKARI-Bench: A Lightweight Benchmark for Comparing Retrieval Architectures and Efficiency Settings under Unified Conditions
HAKARI-Bench reconstructs 35 benchmarks into 551 tasks across 43 languages, reproducing full MTEB, MMTEB, and BEIR rankings with Spearman correlation above 0.97 while supporting efficiency variant comparisons.
-
Compute-Budgeted Exploitability Evidence Graphs for Prospective Vulnerability Triage
Budgeted selection from temporal evidence graphs raises leakage-safe prospective recall@50 for CVE exploitability from 0.010 to 0.026 on 12012 CVEs, with two documents capturing most value and standard splits inflating performance by 8.5x.
-
Non-negative Elastic Net Decoding for Information Retrieval
NNN decoding selects documents via non-negative elastic net reconstruction of the query embedding, with a theorem showing it strictly dominates dense retrieval on correlated corpora and experiments showing gains over inner-product baselines.
-
MetaSyn: A Benchmark for LLM Agents on Meta-Analysis Articles from Nature Portfolio
MetaSyn is a stage-level benchmark of 442 meta-analyses showing LLM agents retrieve up to 90.9% of eligible studies but include at most 52.7% in their final reports.
-
Fast LLM-Based Semantic Filtering: From a Unified Framework to an Adaptive Two-Phase Method
An adaptive two-phase semantic filter using clustering then a hybrid proxy trained on LLM confidence achieves 1.6-2.0x speedup over prior methods at 90% accuracy on 10K document corpora.
-
Re-Ranking Through an Attribution Lens for Citation Quality in Legal QA
Re-ranking retrieval candidates via a cross-encoder trained on continuous perturbation-based attribution scores improves citation faithfulness and gold-answer alignment in legal QA over semantic similarity.
-
Test-Time Training for Zero-Resource Dense Retrieval Reranking
DART adapts a scoring matrix at inference time via gradient updates on pseudo-labels from top/bottom documents to gain +2.1% mean NDCG@10 on six BEIR benchmarks with under 10ms added latency.
-
SilentRetrieval: Hijacking Retrieval-Augmented Generation via Semantically-Preserving Adversarial Data Poisoning
SilentRetrieval is a data poisoning attack achieving 84.6% HR@10 and 57.5% ASR-LLM on Natural Questions via coordinated beam search and trigger fusion while preserving document fluency.
-
Layer-wise Token Compression for Efficient Document Reranking
Layer-wise Token Compression applies adaptive token pooling at middle transformer layers for cross-encoder rerankers, preserving MS MARCO ranking quality while raising QPS up to 25% on passages and 116% on documents, with added gains on listwise LLM rerankers and a regularizer effect for long inputs
-
Very Efficient Listwise Multimodal Reranking for Long Documents
ZipRerank delivers state-of-the-art multimodal listwise reranking accuracy for long documents at up to 10x lower latency via early interaction and single-pass scoring.
-
Hypothesis-Driven Deep Research with Large Language Models: A Structured Methodology for Automated Knowledge Discovery
HDRI is a six-principle eight-stage framework for hypothesis-organized LLM research featuring gap-driven iteration, traceable fact reasoning, and subject locking, realized in INFOMINER with reported gains in fact density and completeness.
-
Prism-Reranker: Beyond Relevance Scoring -- Jointly Producing Contributions and Evidence for Agentic Retrieval
Prism-Reranker models output relevance, contribution statements, and evidence passages to support agentic retrieval beyond scalar scoring.
-
Bayesian Active Learning with Gaussian Processes Guided by LLM Relevance Scoring for Dense Passage Retrieval
BAGEL is a Bayesian active learning framework that uses Gaussian Processes to propagate LLM relevance signals across embedding space and guide global exploration, outperforming standard LLM reranking under identical budgets on four retrieval benchmarks.
-
KIRA: Knowledge-Intensive Image Retrieval and Reasoning Architecture for Specialized Visual Domains
KIRA is a unified architecture for visual RAG that reports 0.97 retrieval precision, 1.0 grounding, and 0.707 domain correctness across medical, circuit, satellite, and histopathology domains via hierarchical chunking, dual-path retrieval, and evidence-conditioned generation.
-
Scaling Laws for Cross-Encoder Reranking
Cross-encoder reranker performance scales predictably via power laws with model size and training exposure, allowing accurate forecasts for 400M and 1B models and data-heavy compute allocation.
-
SPIRE: Structure-Preserving Interpretable Retrieval of Evidence
SPIRE presents a tree-structured retrieval method using subdocuments, paths, and dual contextualization that produces higher-quality and more diverse citations than passage-based baselines on HTML QA benchmarks.
-
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
RAG models set new state-of-the-art results on open-domain QA by retrieving Wikipedia passages and conditioning a generative model on them, while also producing more factual text than parametric baselines.
-
Inject or Navigate? Token-Efficient Retrieval for LLM Analysis of Transactional Legal Documents
Structured retrieval (NAVEMBED/NAVINDEX) matches full-corpus injection on legal multi-document QA while cutting tokens 1.6–30×, with a closed-form cache-crossover cost rule.
-
ContextNest: Verifiable Context Governance for Autonomous AI Agent
Governed selection over versioned, hash-chained knowledge vaults Pareto-dominates BM25 on a stale-version attack and is fully deterministic where dense+HNSW retrieval is not.
-
Relevance Is Not Permission: Warranted Attention for Value Contributions
Warrant adds a query-item permission gate g_ij to attention value terms, improving primary metrics in 27 of 32 comparisons across CTDG, MTPP, RAG, STPP, and TKG tasks.
-
AB-RAG: Adaptive Budgeted Retrieval-Augmented Generation for Reliable Question Answering
AB-RAG adaptively budgets retrieval in RAG by combining three confidence signals to decide when to stop or fetch more evidence, separating correct from incorrect answers at 57.6% vs 0% exact match on a factoid dataset.
-
KaLM-Reranker-V1: Fast but Not Late Interaction for Compressed Document Reranking
KaLM-Reranker-V1 uses encoder–decoder FBNL with Matryoshka pooling to match Qwen3-class reranking quality at substantially lower online cost.
-
HistoRAG: Embedding Historical Methodology in Retrieval-Augmented Generation Through Critical Technical Practice
HistoRAG embeds historiographical principles into RAG via temporal windowing, decoupled retrieval, and contestable LLM relevance judgments, evaluated on 102k Der Spiegel articles from 1950-1979.
-
CoDeR: Local Constraint-Compatible Retrieval Beyond Semantic Similarity
CoDeR augments standard topical dense retrieval with a bi-encoder compatibility scorer trained via contrastive lexical-polarity supervision to reduce early exposure to constraint-violating documents.
-
STORM: Stepwise Token Optimization with Reward-Guided Beam Search
STORM trains lexical query rewriters via reward-guided beam search that converts retrieval metrics into stepwise token signals, enabling 0.6B-8B models to rival dense retrievers on TREC, BEIR and MIRACL without index changes.
-
Membrane: A Self-Evolving Contrastive Safety Memory for LLM Agent Defense
A contrastive memory system evolves without retraining to defend LLM agents against jailbreaks, achieving top F1 scores and low benign refusal on HarmBench and AgentHarm benchmarks.
-
Derivation Prompting: A Logic-Based Method for Improving Retrieval-Augmented Generation
Derivation Prompting constructs logic-based derivation trees in RAG generation to improve interpretability and reduce unacceptable answers compared to standard RAG or long-context methods in a case study.
-
PRISM: Pareto-Efficient Retrieval over Intent-Aware Structured Memory for Long-Horizon Agents
PRISM is a new inference-time retrieval system that achieves higher accuracy than baselines on long-horizon agent tasks while using an order of magnitude less context by combining hierarchical graph search, intent-based costing, compression, and adaptive routing over structured memory.
-
Escaping the Diversity Trap in Robotic Manipulation via Anchor-Centric Adaptation
Anchor-Centric Adaptation escapes the diversity trap by prioritizing repeated demonstrations at core anchors over broad coverage, yielding higher success rates under fixed data budgets in robotic manipulation.
-
Storage Is Not Memory: A Retrieval-Centered Architecture for Agent Recall
True Memory is a verbatim-event retrieval pipeline running on a single SQLite file that reaches 93% accuracy on LoCoMo multi-session questions, outperforming Mem0, Supermemory, Zep, and matching or exceeding EverMemOS and Hindsight on other long-context benchmarks.
-
Interactive Multi-Turn Retrieval for Health Videos
DATR combines coarse CLIP-based retrieval with multi-turn query fusion and cross-encoder re-ranking to improve health video retrieval, supported by the new MHVRC corpus.
-
Verbal-R3: Verbal Reranker as the Missing Bridge between Retrieval and Reasoning
Verbal-R3 uses a verbal reranker to generate analytic narratives that guide retrieval and reasoning in LLMs, achieving SOTA results on complex QA benchmarks.
-
A Replicability Study of XTR
XTR training does not improve retrieval effectiveness over ColBERT but enhances IVF engine efficiency by flattening token scores to produce more discriminative centroids.
-
Onyx: Cost-Efficient Disk-Oblivious ANN Search
Onyx inverts ANN-ORAM optimization priorities with a compact pruning representation and locality-aware shallow tree to deliver 1.7-9.9x lower cost and 2.3-12.3x lower latency for disk-oblivious ANN search.
-
The Effect of Document Selection on Query-focused Text Analysis
Semantic and hybrid document retrieval methods provide reliable, efficient selection for query-focused text analyses like LDA and BERTopic, outperforming random or keyword-only approaches.
-
Entities as Retrieval Signals: A Systematic Study of Coverage, Supervision, and Evaluation in Entity-Oriented Ranking
Entity-oriented rerankers show large gains in conditional evaluation but not open-world evaluation on Robust04; the paper attributes this to a coverage–discrimination tradeoff in the entity channel, not to model architecture.
-
Beyond Single-Score Ranking: Facet-Aware Reranking for Controllable Diversity in Paper Recommendation
SciFACE improves facet-specific paper ranking NDCG scores by training separate cross-encoders for Background and Method similarity on 5,891 GPT-4o-mini labeled pairs, outperforming SPECTER by up to 31 points.
-
Web Retrieval-Aware Chunking (W-RAC) for Efficient and Cost-Effective Retrieval-Augmented Generation Systems
W-RAC decouples extraction from semantic planning via structured units and LLM grouping to match traditional retrieval performance at roughly 10x lower LLM token cost.
-
ProRank: Prompt Warmup via Reinforcement Learning for Small Language Models Reranking
ProRank uses RL-based prompt warmup and fine-grained scoring to train small language models that surpass LLM rerankers on BEIR.
-
RankZephyr: Effective and Robust Zero-Shot Listwise Reranking is a Breeze!
RankZephyr is a new open-source LLM that closes the effectiveness gap with GPT-4 for zero-shot listwise reranking while showing robustness to input ordering and document count.
-
Atlas: Few-shot Learning with Retrieval Augmented Language Models
Atlas reaches over 42% accuracy on Natural Questions with only 64 examples, outperforming a 540B-parameter model by 3% with 50x fewer parameters.
-
Unsupervised Dense Information Retrieval with Contrastive Learning
Contrastive learning trains unsupervised dense retrievers that beat BM25 on most BEIR datasets and support cross-lingual retrieval across scripts.
-
SkillSelect-Serve: QoS-Aware Budgeted Skill Service Recommendation for LLM Agents
A QoS-aware constrained packer restores 100% skill deliverability under token/risk/tool budgets at ~1.14 hit-rate points cost over unconstrained Top-5 on 35k skills.
-
Multi-Agent Routing as Set-Valued Prediction: A WildChat Benchmark and Cost-Aware Evaluation
On a rebalanced WildChat 12-agent benchmark, fine-tuned encoders best match multi-agent gold sets, while Weighted Agent Routing improves cost-aware utility on strong scorers.
-
Adaptive Re-Ranking
Adaptive Re-Ranking trains a classifier to route queries to BM25, MiniLM-L6-v2, or BGE-v2-m3 based on a utility label, yielding 1.15-53x lower median latency and competitive nDCG@10 versus always using the heaviest model.