Fine-tuning BERT for query-passage relevance classification achieves state-of-the-art results on TREC-CAR and MS MARCO, with a 27% relative gain in MRR@10 over prior methods.
hub Tool reference
MS MARCO: A Human Generated MAchine Reading COmprehension Dataset
Tool reference. 93% of classified Pith citations use this work as a method, library, or software dependency, not as a substantive claim.
68
Pith papers citing it
Method reference
93% of classified citations
abstract
We introduce a large scale MAchine Reading COmprehension dataset, which we name MS MARCO. The dataset comprises of 1,010,916 anonymized questions---sampled from Bing's search query logs---each with a human generated answer and 182,669 completely human rewritten generated answers. In addition, the dataset contains 8,841,823 passages---extracted from 3,563,535 web documents retrieved by Bing---that provide the information necessary for curating the natural language answers. A question in the MS MARCO dataset may have multiple answers or no answers at all. Using this dataset, we propose three different tasks with varying levels of difficulty: (i) predict if a question is answerable given a set of context passages, and extract and synthesize the answer as a human would (ii) generate a well-formed answer (if possible) based on the context passages that can be understood with the question and passage context, and finally (iii) rank a set of retrieved passages given a question. The size of the dataset and the fact that the questions are derived from real user search queries distinguishes MS MARCO from other well-known publicly available datasets for machine reading comprehension and question-answering. We believe that the scale and the real-world nature of this dataset makes it attractive for benchmarking machine reading comprehension and question-answering models.
hub tools
citation-role summary
dataset 11
method 2
background 1
citation-polarity summary
claims ledger
- abstract We introduce a large scale MAchine Reading COmprehension dataset, which we name MS MARCO. The dataset comprises of 1,010,916 anonymized questions---sampled from Bing's search query logs---each with a human generated answer and 182,669 completely human rewritten generated answers. In addition, the dataset contains 8,841,823 passages---extracted from 3,563,535 web documents retrieved by Bing---that provide the information necessary for curating the natural language answers. A question in the MS MARCO dataset may have multiple answers or no answers at all. Using this dataset, we propose three dif
co-cited works
representative citing papers
TriviaQA is a new large-scale dataset for reading comprehension that features complex compositional questions, high lexical variability, and cross-sentence reasoning requirements, where current baselines reach only 40% while humans reach 80%.
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
On-policy distillation has an extrapolation cliff at closed-form lambda*(p,b,c) set by teacher modal probability, warm-start mass, and clip strength, past which training shifts from format-preserving to format-collapsing.
BRIGHT-Pro and RTriever-Synth advance reasoning-intensive retrieval by adding multi-aspect evidence evaluation and aspect-decomposed synthetic training, with the fine-tuned RTriever-4B showing gains over its base model.
Modern text encoders resist second-order collapse under mean pooling because token embeddings concentrate tightly within texts, and this resistance correlates with stronger downstream performance.
UnIte selects target-domain documents for pseudo-query generation by filtering high aleatoric uncertainty and prioritizing high epistemic uncertainty, yielding +2.45 to +3.49 nDCG@10 gains on BEIR with ~4k samples.
A product-key parametric memory head with selective sparse updates mitigates catastrophic forgetting in generative retrieval models during sequential addition of new documents.
LLM-based dense retrievers generalize better when instruction-tuned but pay a specialization tax when optimized for reasoning; they resist typos and corpus poisoning better than encoder-only baselines yet remain vulnerable to semantic perturbations, with larger models and certain embedding geometry,
A single query-specific poisoned document, built by extracting and iteratively refining an adversarial chain-of-thought, can substantially degrade reasoning accuracy in retrieval-augmented LLM systems.
Injecting a few malicious vectors near the centroid exploits centrality-driven hubness in high-dimensional embeddings, causing them to dominate top-k retrievals in up to 99.85% of cases.
Spectral Tempering derives an adaptive scaling factor γ(k) from the embedding eigenspectrum via local SNR analysis and knee-point normalization to achieve near-optimal compression without training or validation.
ScrapeGraphAI-100k releases 93,695 real telemetry examples pairing web page content with prompts, schemas, and LLM responses to support training and benchmarking of schema-constrained generation.
SQuTR aggregates 37k queries from six text retrieval datasets, synthesizes speech from 200 speakers, adds 17 noise categories at varying SNR, and shows that even large retrieval models degrade sharply under extreme acoustic noise.
A survey that defines Compound AI Systems, proposes a multi-dimensional taxonomy based on component roles and orchestration strategies, reviews four foundational paradigms, and identifies key challenges for future research.
GAIA benchmark shows humans at 92% accuracy on simple real-world questions far outperform current AI systems at 15%, proposing this gap as a key milestone for general AI.
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.
BoolQ introduces naturally occurring yes/no questions as a challenging benchmark where BERT fine-tuned on MultiNLI reaches 80.4% accuracy against 90% human performance.
SPECTRA generates reproducible synthetic IR corpora up to 60,000 documents with controllable distractors, long-tail vocabulary, and graded relevance labels via a single-process Python prototype.
A Llama-based model trained on serialized user stories unifies item, carousel, and search ranking and outperforms specialist baselines offline while improving some online metrics and reducing latency.
NasZip delivers up to 8.4x speedup over CPU baselines and 1.69x over prior NDP accelerators for ANNS by combining near-data processing with statistics-based PCA early exiting, dynamic-float encoding, and data-aware neighbor mapping.
MAFIG is a multi-agent framework that uses LLM agents and evaluators to generate reading comprehension items with significantly higher adherence to specified feature constraints than single-agent baselines.
A dual hierarchical RL framework with two agents coordinates high-level dialogue strategy and low-level question generation to emulate judicial questioning and extract key information from Supreme Court arguments, outperforming baselines.
EnterpriseRAG-Bench supplies a synthetic corpus of 500k documents across Slack, Gmail, Linear, Google Drive, HubSpot, Fireflies, GitHub, Jira and Confluence together with 500 questions spanning single-document lookup to conflict resolution and missing-information detection.
citing papers explorer
-
Passage Re-ranking with BERT
Fine-tuning BERT for query-passage relevance classification achieves state-of-the-art results on TREC-CAR and MS MARCO, with a 27% relative gain in MRR@10 over prior methods.
-
TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension
TriviaQA is a new large-scale dataset for reading comprehension that features complex compositional questions, high lexical variability, and cross-sentence reasoning requirements, where current baselines reach only 40% while humans reach 80%.
-
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
-
The Extrapolation Cliff in On-Policy Distillation of Near-Deterministic Structured Outputs
On-policy distillation has an extrapolation cliff at closed-form lambda*(p,b,c) set by teacher modal probability, warm-start mass, and clip strength, past which training shifts from format-preserving to format-collapsing.
-
Rethinking Reasoning-Intensive Retrieval: Evaluating and Advancing Retrievers in Agentic Search Systems
BRIGHT-Pro and RTriever-Synth advance reasoning-intensive retrieval by adding multi-aspect evidence evaluation and aspect-decomposed synthetic training, with the fine-tuned RTriever-4B showing gains over its base model.
-
Why Mean Pooling Works: Quantifying Second-Order Collapse in Text Embeddings
Modern text encoders resist second-order collapse under mean pooling because token embeddings concentrate tightly within texts, and this resistance correlates with stronger downstream performance.
-
UnIte: Uncertainty-based Iterative Document Sampling for Domain Adaptation in Information Retrieval
UnIte selects target-domain documents for pseudo-query generation by filtering high aleatoric uncertainty and prioritizing high epistemic uncertainty, yielding +2.45 to +3.49 nDCG@10 gains on BEIR with ~4k samples.
-
A Parametric Memory Head for Continual Generative Retrieval
A product-key parametric memory head with selective sparse updates mitigates catastrophic forgetting in generative retrieval models during sequential addition of new documents.
-
On the Robustness of LLM-Based Dense Retrievers: A Systematic Analysis of Generalizability and Stability
LLM-based dense retrievers generalize better when instruction-tuned but pay a specialization tax when optimized for reasoning; they resist typos and corpus poisoning better than encoder-only baselines yet remain vulnerable to semantic perturbations, with larger models and certain embedding geometry,
-
AdversarialCoT: Single-Document Retrieval Poisoning for LLM Reasoning
A single query-specific poisoned document, built by extracting and iteratively refining an adversarial chain-of-thought, can substantially degrade reasoning accuracy in retrieval-augmented LLM systems.
-
Can You Trust the Vectors in Your Vector Database? Black-Hole Attack from Embedding Space Defects
Injecting a few malicious vectors near the centroid exploits centrality-driven hubness in high-dimensional embeddings, causing them to dominate top-k retrievals in up to 99.85% of cases.
-
Spectral Tempering for Embedding Compression in Dense Passage Retrieval
Spectral Tempering derives an adaptive scaling factor γ(k) from the embedding eigenspectrum via local SNR analysis and knee-point normalization to achieve near-optimal compression without training or validation.
-
ScrapeGraphAI-100k: Dataset for Schema-Constrained LLM Generation
ScrapeGraphAI-100k releases 93,695 real telemetry examples pairing web page content with prompts, schemas, and LLM responses to support training and benchmarking of schema-constrained generation.
-
SQuTR: A Robustness Benchmark for Spoken Query to Text Retrieval under Acoustic Noise
SQuTR aggregates 37k queries from six text retrieval datasets, synthesizes speech from 200 speakers, adds 17 noise categories at varying SNR, and shows that even large retrieval models degrade sharply under extreme acoustic noise.
-
From Standalone LLMs to Integrated Intelligence: A Survey of Compound Al Systems
A survey that defines Compound AI Systems, proposes a multi-dimensional taxonomy based on component roles and orchestration strategies, reviews four foundational paradigms, and identifies key challenges for future research.
-
GAIA: a benchmark for General AI Assistants
GAIA benchmark shows humans at 92% accuracy on simple real-world questions far outperform current AI systems at 15%, proposing this gap as a key milestone for general AI.
-
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.
-
BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions
BoolQ introduces naturally occurring yes/no questions as a challenging benchmark where BERT fine-tuned on MultiNLI reaches 80.4% accuracy against 90% human performance.
-
SPECTRA: Synthetic IR Test Collections with Relevance Oracles and Controlled Distractor Diagnostics
SPECTRA generates reproducible synthetic IR corpora up to 60,000 documents with controllable distractors, long-tail vocabulary, and graded relevance labels via a single-process Python prototype.
-
TubiFM: Unified Item, Carousel, and Search Ranking for Streaming Discovery
A Llama-based model trained on serialized user stories unifies item, carousel, and search ranking and outperforms specialist baselines offline while improving some online metrics and reducing latency.
-
NasZip: Software and Hardware Co-Design to Accelerate Approximate Nearest Neighbor Search with DIMM-Based Near-Data Processing
NasZip delivers up to 8.4x speedup over CPU baselines and 1.69x over prior NDP accelerators for ANNS by combining near-data processing with statistics-based PCA early exiting, dynamic-float encoding, and data-aware neighbor mapping.
-
A Multi-Agent Framework for Feature-Constrained Difficulty Control in Reading Comprehension Item Generation
MAFIG is a multi-agent framework that uses LLM agents and evaluators to generate reading comprehension items with significantly higher adherence to specified feature constraints than single-agent baselines.
-
Dual Hierarchical Dialogue Policy Learning for Legal Inquisitive Conversational Agents
A dual hierarchical RL framework with two agents coordinates high-level dialogue strategy and low-level question generation to emulate judicial questioning and extract key information from Supreme Court arguments, outperforming baselines.
-
EnterpriseRAG-Bench: A RAG Benchmark for Company Internal Knowledge
EnterpriseRAG-Bench supplies a synthetic corpus of 500k documents across Slack, Gmail, Linear, Google Drive, HubSpot, Fireflies, GitHub, Jira and Confluence together with 500 questions spanning single-document lookup to conflict resolution and missing-information detection.
-
Reproducing Complex Set-Compositional Information Retrieval
Neural retrievers that double BM25 performance on QUEST collapse below 0.02 Recall@100 on the new LIMIT+ benchmark while lexical methods reach 0.96, with all methods degrading as compositional depth increases.
-
NuggetIndex: Governed Atomic Retrieval for Maintainable RAG
NuggetIndex manages atomic nuggets with temporal validity and lifecycle metadata to filter outdated information before ranking, yielding 42% higher nugget recall, 9pp better temporal correctness, and 55% fewer conflicts than passage or unmanaged proposition baselines.
-
RAQG-QPP: Query Performance Prediction with Retrieved Query Variants and Retrieval Augmented Query Generation
Retrieved query variants from logs combined with LLM-augmented generation improve unsupervised QPP accuracy by up to 30% for neural rankers on TREC DL'19 and DL'20.
-
JigsawRL: Assembling RL Pipelines for Efficient LLM Post-Training
JigsawRL achieves up to 1.85x higher throughput in LLM RL pipelines via pipeline multiplexing, sub-stage graphs, and look-ahead scheduling compared to prior systems.
-
ORPHEAS: A Cross-Lingual Greek-English Embedding Model for Retrieval-Augmented Generation
ORPHEAS, a Greek-English embedding model created with knowledge graph fine-tuning, outperforms state-of-the-art multilingual models on monolingual and cross-lingual retrieval benchmarks.
-
Stochastic KV Routing: Enabling Adaptive Depth-Wise Cache Sharing
Stochastic training with random cross-layer KV attention enables depth-wise cache sharing in transformers, cutting memory footprint while preserving or improving performance.
-
A Voronoi Cell Formulation for Principled Token Pruning in Late-Interaction Retrieval Models
A Voronoi cell estimation framework in embedding space enables principled token pruning for late-interaction models, reducing index size while retaining retrieval quality.
-
MSA: Memory Sparse Attention for Efficient End-to-End Memory Model Scaling to 100M Tokens
MSA is an end-to-end trainable memory model using sparse attention and document-wise RoPE that scales to 100M tokens with linear complexity and less than 9% degradation.
-
Towards Efficient and Generalizable Retrieval: Adaptive Semantic Quantization and Residual Knowledge Transfer
SA²CRQ uses sequential adaptive residual quantization based on path entropy plus anchored curriculum regularization from head items to improve both efficiency and cold-start performance in generative retrieval.
-
When LLM Judges Inflate Scores: Exploring Overrating in Relevance Assessment
LLMs consistently overrate relevance of inadequate passages in IR evaluations due to biases toward length and lexical features rather than true content match.
-
LLM Prompt Duel Optimizer: Efficient Label-Free Prompt Optimization
Prompt Duel Optimizer uses dueling bandits and LLM-as-judge pairwise feedback with Double Thompson Sampling and top-performer mutation to find stronger prompts than label-free baselines on BBH and MS MARCO under limited comparison budgets.
-
LEAF: Knowledge Distillation of Text Embedding Models with Teacher-Aligned Representations
LEAF distills teacher-aligned student embedding models that achieve new SOTA results on BEIR and MTEB for their size class while requiring only modest data and compute.
-
Learning from Natural Language Feedback for Personalized Question Answering
VAC replaces scalar rewards with natural language feedback in an alternating training loop between a feedback model and a policy model, yielding better personalized QA on the LaMP-QA benchmark.
-
Should We Still Pretrain Encoders with Masked Language Modeling?
Controlled ablations of 38 models find MLM superior to CLM on representation benchmarks while CLM offers better data efficiency and stability; a biphasic CLM-then-MLM schedule is optimal under fixed compute and improves when initialized from pretrained CLM models.
-
RankFlow: A Multi-Role Collaborative Reranking Workflow Utilizing Large Language Models
RankFlow deploys four LLM roles in sequence to rewrite queries, generate pseudo-answers, summarize passages, and rerank candidates, outperforming prior methods on TREC-DL, BEIR, and NovelEval.
-
NV-Embed: Improved Techniques for Training LLMs as Generalist Embedding Models
NV-Embed achieves first place on the MTEB leaderboard across 56 tasks by combining a latent attention layer, causal-mask removal, two-stage contrastive training, and data curation for LLM-based embedding models.
-
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.
-
REPLUG: Retrieval-Augmented Black-Box Language Models
REPLUG improves frozen black-box LMs by prepending LM-supervised retrieved documents, delivering 6.3% better language modeling on GPT-3 and 5.1% better five-shot MMLU on Codex.
-
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.
-
Text and Code Embeddings by Contrastive Pre-Training
Contrastive pre-training on unsupervised data at scale creates text and code embeddings that set new state-of-the-art results on classification and semantic search benchmarks.
-
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.
-
NAVIS: Concurrent Search and Update with Low Position-Seeking Overhead in On-SSD Graph-Based Vector Search
NAVIS improves concurrent search and update throughput in on-SSD graph vector search by up to 2.74x for insertions and 1.37x for searches through reduced position-seeking overhead.
-
Efficient Listwise Reranking with Compressed Document Representations
RRK compresses documents to multi-token embeddings for efficient listwise reranking, enabling an 8B model to achieve 3x-18x speedups over smaller models with comparable or better effectiveness.
-
RefineRAG: Word-Level Poisoning Attacks via Retriever-Guided Text Refinement
RefineRAG achieves 90% attack success on NQ by generating toxic seeds then optimizing them via retriever-in-the-loop word refinement, outperforming prior methods on effectiveness and naturalness.
-
GaiaFlow: Semantic-Guided Diffusion Tuning for Carbon-Frugal Search
GaiaFlow combines semantic-guided diffusion tuning with early-exit and quantization methods to lower carbon emissions in neural information retrieval while maintaining competitive effectiveness.
-
Do Activation Verbalization Methods Convey Privileged Information?
Activation verbalization methods for LLMs largely reflect the verbalizer model's parametric knowledge rather than privileged information from the target model's activations.