Keping Bi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Keping Bi has a total of 62 co-authors. Here are the top co-authors by number of shared publications:
- Jiafeng Guo
School of Computer Science and Technology, University of Chinese Academy of Sciences, Beijing, China
34 shared publications - Xueqi Cheng
Chinese Academy of Sciences University of Chinese Academy of Sciences Beijing China
18 shared publications - Hengran Zhang
State Key Laboratory of AI Safety, Beijing, China, Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China, and University of Chinese Academy of Sciences, Beijing, China
9 shared publications - Dawei Yin
Baidu Inc., Beijing, China
7 shared publications - Xiaojie Sun
State Key Laboratory of AI Safety, Beijing, China, Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China, and University of Chinese Academy of Sciences, Beijing, China
6 shared publications Xueqi Cheng
State Key Laboratory of AI Safety Institute of Computing Technology Chinese Academy of Sciences
5 shared publications
Latest Publications
- Open Access
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 5830-5841
Can LLM Annotations Replace User Clicks for Learning to Rank?
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 5830-5841
Open AccessCan LLM Annotations Replace User Clicks for Learning to Rank?
- Open Access
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 6126-6137
LLM-Specific Utility for Retrieval-Augmented Generation
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 6126-6137
Open AccessLLM-Specific Utility for Retrieval-Augmented Generation
- Open Access
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 8064-8069
Unleashing the Power of LLMs in Dense Retrieval with Query Likelihood Modeling
- 6 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 8064-8069
Open AccessUnleashing the Power of LLMs in Dense Retrieval with Query Likelihood Modeling
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4766-4777
Training Dense Retrievers with Multiple Positive Passages
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4766-4777
Open AccessTraining Dense Retrievers with Multiple Positive Passages
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 156-166
Reconstructing Content with Collaborative Attention for Universal Multimodal Representation Learning
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 156-166
Open AccessReconstructing Content with Collaborative Attention for Universal Multimodal Representation Learning
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 5357-5361
Beyond Relevance: Utility-Centric Retrieval in the LLM Era
Information retrieval systems have traditionally optimized for topical relevance—the degree to which retrieved documents match a query. However, relevance only approximates a deeper goal: utility, namely, whether retrieved information helps accomplish a user's underlying task. The emergence of retrieval-augmented generation (RAG) fundamentally changes this paradigm. Retrieved documents are no longer consumed directly by users but instead serve as evidence for large language models (LLMs) that produce answers. As a result, retrieval effectiveness must be evaluated by its contribution to generation quality rather than by relevance-based ranking metrics alone. This tutorial argues that retrieval objectives are evolving from relevance-centric optimization toward LLM-centric utility. We present a unified framework covering LLM-agnostic versus LLM-specific utility, context-independent versus context-dependent utility, and the connection with LLM information needs and agentic RAG. By synthesizing recent advances, the tutorial provides conceptual foundations and practical guidance for designing retrieval systems aligned with the requirements of LLM-based information access.
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 5357-5361
Open AccessBeyond Relevance: Utility-Centric Retrieval in the LLM Era
Information retrieval systems have traditionally optimized for topical relevance—the degree to which retrieved documents match a query. However, relevance only approximates a deeper goal: utility, namely, whether retrieved information helps accomplish a user's underlying task. The emergence of retrieval-augmented generation (RAG) fundamentally changes this paradigm. Retrieved documents are no longer consumed directly by users but instead serve as evidence for large language models (LLMs) that produce answers. As a result, retrieval effectiveness must be evaluated by its contribution to generation quality rather than by relevance-based ranking metrics alone. This tutorial argues that retrieval objectives are evolving from relevance-centric optimization toward LLM-centric utility. We present a unified framework covering LLM-agnostic versus LLM-specific utility, context-independent versus context-dependent utility, and the connection with LLM information needs and agentic RAG. By synthesizing recent advances, the tutorial provides conceptual foundations and practical guidance for designing retrieval systems aligned with the requirements of LLM-based information access.
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 223-233
Attention Grounded Enhancement for Visual Document Retrieval
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 223-233
Open AccessAttention Grounded Enhancement for Visual Document Retrieval
- Open Access
- 21 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 238-248
How Do LLM-Generated Texts Impact Term-Based Retrieval Models?
- 21 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 238-248
Open AccessHow Do LLM-Generated Texts Impact Term-Based Retrieval Models?
- Open Access
Training Dense Retrievers with Multiple Positive Passages
Modern knowledge-intensive systems, such as retrieval-augmented generation (RAG), rely on effective retrievers to establish the performance ceiling for downstream modules. However, retriever training has been bottlenecked by sparse, single-positive annotations, which lead to false-negative noise and suboptimal supervision. While the advent of large language models (LLMs) makes it feasible to collect comprehensive multi-positive relevance labels at scale, the optimal strategy for incorporating these dense signals into training remains poorly understood. In this paper, we present a systematic study of multi-positive optimization objectives for retriever training. We unify representative objectives, including Joint Likelihood (JointLH), Summed Marginal Likelihood (SumMargLH), and Log-Sum-Exp Pairwise (LSEPair) loss, under a shared contrastive learning framework. Our theoretical analysis characterizes their distinct gradient behaviors, revealing how each allocates probability mass across positive document sets. Empirically, we conduct extensive evaluations on Natural Questions, MS MARCO, and the BEIR benchmark across two realistic regimes: homogeneous LLM-annotated data and heterogeneous mixtures of human and LLM labels. Our results show that LSEPair consistently achieves superior robustness and performance across settings, while JointLH and SumMargLH exhibit high sensitivity to the quality of positives. Furthermore, we find that the simple strategy of random sampling (Rand1LH) serves as a reliable baseline. By aligning theoretical insights with empirical findings, we provide practical design principles for leveraging dense, LLM-augmented supervision to enhance retriever effectiveness.
Open AccessTraining Dense Retrievers with Multiple Positive Passages
Modern knowledge-intensive systems, such as retrieval-augmented generation (RAG), rely on effective retrievers to establish the performance ceiling for downstream modules. However, retriever training has been bottlenecked by sparse, single-positive annotations, which lead to false-negative noise and suboptimal supervision. While the advent of large language models (LLMs) makes it feasible to collect comprehensive multi-positive relevance labels at scale, the optimal strategy for incorporating these dense signals into training remains poorly understood. In this paper, we present a systematic study of multi-positive optimization objectives for retriever training. We unify representative objectives, including Joint Likelihood (JointLH), Summed Marginal Likelihood (SumMargLH), and Log-Sum-Exp Pairwise (LSEPair) loss, under a shared contrastive learning framework. Our theoretical analysis characterizes their distinct gradient behaviors, revealing how each allocates probability mass across positive document sets. Empirically, we conduct extensive evaluations on Natural Questions, MS MARCO, and the BEIR benchmark across two realistic regimes: homogeneous LLM-annotated data and heterogeneous mixtures of human and LLM labels. Our results show that LSEPair consistently achieves superior robustness and performance across settings, while JointLH and SumMargLH exhibit high sensitivity to the quality of positives. Furthermore, we find that the simple strategy of random sampling (Rand1LH) serves as a reliable baseline. By aligning theoretical insights with empirical findings, we provide practical design principles for leveraging dense, LLM-augmented supervision to enhance retriever effectiveness.
- Open Access
Bagging-Based Model Merging for Robust General Text Embeddings
General-purpose text embedding models underpin a wide range of NLP and information retrieval applications, and are typically trained on large-scale multi-task corpora to encourage broad generalization. However, it remains unclear how different multi-task training strategies compare in practice, and how to efficiently adapt embedding models as new domains and data types continually emerge. In this work, we present a systematic study of multi-task training for text embeddings from two perspectives: data scheduling and model merging. We compare batch-level shuffling, sequential training variants, two-stage training, and multiple merging granularities, and find that simple batch-level shuffling consistently yields the strongest overall performance, suggesting that task conflicts are limited and training datasets are largely complementary. Despite its effectiveness, batch-level shuffling exhibits two practical limitations: suboptimal out-of-domain (OOD) generalization and poor suitability for incremental learning due to expensive full retraining. To address these issues, we propose Bagging-based rObust mOdel Merging (BOOM), which trains multiple embedding models on sampled subsets and merges them into a single model, improving robustness while retaining single-model inference efficiency. Moreover, BOOM naturally supports efficient incremental updates by training lightweight update models on new data with a small historical subset and merging them into the existing model. Experiments across diverse embedding benchmarks demonstrate that BOOM consistently improves both in-domain and OOD performance over full-corpus batch-level shuffling, while substantially reducing training cost in incremental learning settings.
Open AccessBagging-Based Model Merging for Robust General Text Embeddings
General-purpose text embedding models underpin a wide range of NLP and information retrieval applications, and are typically trained on large-scale multi-task corpora to encourage broad generalization. However, it remains unclear how different multi-task training strategies compare in practice, and how to efficiently adapt embedding models as new domains and data types continually emerge. In this work, we present a systematic study of multi-task training for text embeddings from two perspectives: data scheduling and model merging. We compare batch-level shuffling, sequential training variants, two-stage training, and multiple merging granularities, and find that simple batch-level shuffling consistently yields the strongest overall performance, suggesting that task conflicts are limited and training datasets are largely complementary. Despite its effectiveness, batch-level shuffling exhibits two practical limitations: suboptimal out-of-domain (OOD) generalization and poor suitability for incremental learning due to expensive full retraining. To address these issues, we propose Bagging-based rObust mOdel Merging (BOOM), which trains multiple embedding models on sampled subsets and merges them into a single model, improving robustness while retaining single-model inference efficiency. Moreover, BOOM naturally supports efficient incremental updates by training lightweight update models on new data with a small historical subset and merging them into the existing model. Experiments across diverse embedding benchmarks demonstrate that BOOM consistently improves both in-domain and OOD performance over full-corpus batch-level shuffling, while substantially reducing training cost in incremental learning settings.