Jamie Callan
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
Jamie Callan has a total of 71 co-authors. Here are the top co-authors by number of shared publications:
- Luyu Gao
Language Technologies Institute Carnegie Mellon University
10 shared publications - Bruno Martins
Instituto Superior Técnico and INESC-ID, Lisbon, Portugal
4 shared publications - Jimmy Lin
University of Waterloo, Waterloo, Canada
4 shared publications - João Magalhães
NOVA LINCS NOVA School of Science and Technology Portugal
4 shared publications - Zhuyun Dai
Carnegie Mellon University
3 shared publications Bhavana Dalvi
School of Computer Science Carnegie Mellon University Pittsburgh 15213 PA
2 shared publications
Latest Publications
- Open Access
- 24 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 34-43
DeepResearchGym: A Free, Transparent, and Reproducible Sandbox for Deep Research
- 24 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 34-43
Open AccessDeepResearchGym: A Free, Transparent, and Reproducible Sandbox for Deep Research
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1404-1415
Agentic Search in the Wild: Intents and Trajectory Dynamics from 14M+ Real Search Requests
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1404-1415
Open AccessAgentic Search in the Wild: Intents and Trajectory Dynamics from 14M+ Real Search Requests
- 25 March 2026
- Published by Springer Nature
- p. 598-613
Less LLM, More Documents: Searching for Improved RAG
- 25 March 2026
- Published by Springer Nature
- p. 598-613
Less LLM, More Documents: Searching for Improved RAG
- Open Access
- 13 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 4061-4065
Tevatron 2.0: Unified Document Retrieval Toolkit across Scale, Language, and Modality
- 13 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 4061-4065
Open AccessTevatron 2.0: Unified Document Retrieval Toolkit across Scale, Language, and Modality
- Open Access
Repository-level Code Search with Neural Retrieval Methods
This paper presents a multi-stage reranking system for repository-level code search, which leverages the vastly available commit histories of large open-source repositories to aid in bug fixing. We define the task of repository-level code search as retrieving the set of files from the current state of a code repository that are most relevant to addressing a user's question or bug. The proposed approach combines BM25-based retrieval over commit messages with neural reranking using CodeBERT to identify the most pertinent files. By learning patterns from diverse repositories and their commit histories, the system can surface relevant files for the task at hand. The system leverages both commit messages and source code for relevance matching, and is evaluated in both normal and oracle settings. Experiments on a new dataset created from 7 popular open-source repositories demonstrate substantial improvements of up to 80% in MAP, MRR and P@1 over the BM25 baseline, across a diverse set of queries, demonstrating the effectiveness this approach. We hope this work aids LLM agents as a tool for better code search and understanding. Our code and results obtained are publicly available.
Open AccessRepository-level Code Search with Neural Retrieval Methods
This paper presents a multi-stage reranking system for repository-level code search, which leverages the vastly available commit histories of large open-source repositories to aid in bug fixing. We define the task of repository-level code search as retrieving the set of files from the current state of a code repository that are most relevant to addressing a user's question or bug. The proposed approach combines BM25-based retrieval over commit messages with neural reranking using CodeBERT to identify the most pertinent files. By learning patterns from diverse repositories and their commit histories, the system can surface relevant files for the task at hand. The system leverages both commit messages and source code for relevance matching, and is evaluated in both normal and oracle settings. Experiments on a new dataset created from 7 popular open-source repositories demonstrate substantial improvements of up to 80% in MAP, MRR and P@1 over the BM25 baseline, across a diverse set of queries, demonstrating the effectiveness this approach. We hope this work aids LLM agents as a tool for better code search and understanding. Our code and results obtained are publicly available.
- Open Access
ACER: Automatic Language Model Context Extension via Retrieval
Long-context modeling is one of the critical capabilities of language AI for digesting and reasoning over complex information pieces. In practice, long-context capabilities are typically built into a pre-trained language model~(LM) through a carefully designed context extension stage, with the goal of producing generalist long-context capabilities. In our preliminary experiments, however, we discovered that the current open-weight generalist long-context models are still lacking in practical long-context processing tasks. While this means perfectly effective long-context modeling demands task-specific data, the cost can be prohibitive. In this paper, we draw inspiration from how humans process a large body of information: a lossy \textbf{retrieval} stage ranks a large set of documents while the reader ends up reading deeply only the top candidates. We build an \textbf{automatic} data synthesis pipeline that mimics this process using short-context LMs. The short-context LMs are further tuned using these self-generated data to obtain task-specific long-context capabilities. Similar to how pre-training learns from imperfect data, we hypothesize and further demonstrate that the short-context model can bootstrap over the synthetic data, outperforming not only long-context generalist models but also the retrieval and read pipeline used to synthesize the training data in real-world tasks such as long-context retrieval augmented generation.
Open AccessACER: Automatic Language Model Context Extension via Retrieval
Long-context modeling is one of the critical capabilities of language AI for digesting and reasoning over complex information pieces. In practice, long-context capabilities are typically built into a pre-trained language model~(LM) through a carefully designed context extension stage, with the goal of producing generalist long-context capabilities. In our preliminary experiments, however, we discovered that the current open-weight generalist long-context models are still lacking in practical long-context processing tasks. While this means perfectly effective long-context modeling demands task-specific data, the cost can be prohibitive. In this paper, we draw inspiration from how humans process a large body of information: a lossy \textbf{retrieval} stage ranks a large set of documents while the reader ends up reading deeply only the top candidates. We build an \textbf{automatic} data synthesis pipeline that mimics this process using short-context LMs. The short-context LMs are further tuned using these self-generated data to obtain task-specific long-context capabilities. Similar to how pre-training learns from imperfect data, we hypothesize and further demonstrate that the short-context model can bootstrap over the synthetic data, outperforming not only long-context generalist models but also the retrieval and read pipeline used to synthesize the training data in real-world tasks such as long-context retrieval augmented generation.
- Open Access
- 10 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 2437-2441
Generalizable Tip-of-the-Tongue Retrieval with LLM Re-ranking
Tip-of-the-Tongue (ToT) retrieval is challenging for search engines because the queries are usually natural-language, verbose, and contain uncertain and inaccurate information. This paper studies the generalization capabilities of existing retrieval methods with ToT queries in multiple domains. We curate a multi-domain dataset and evaluate the effectiveness of recall-oriented first-stage retrieval methods across the different domains, considering in-domain, out-of-domain, and multi-domain training settings. We further explore the use of a Large Language Model (LLM), i.e. GPT-4, for zero-shot re-ranking in various ToT domains, relying solely on the item titles. Results show that multi-domain training enhances recall, and that LLMs are strong zero-shot re-rankers, especially for popular items, outperforming direct GPT-4 prompting without first-stage retrieval. Datasets and code can be found on GitHub https://github.com/LuisPB7/TipTongue
- 10 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 2437-2441
Open AccessGeneralizable Tip-of-the-Tongue Retrieval with LLM Re-ranking
Tip-of-the-Tongue (ToT) retrieval is challenging for search engines because the queries are usually natural-language, verbose, and contain uncertain and inaccurate information. This paper studies the generalization capabilities of existing retrieval methods with ToT queries in multiple domains. We curate a multi-domain dataset and evaluate the effectiveness of recall-oriented first-stage retrieval methods across the different domains, considering in-domain, out-of-domain, and multi-domain training settings. We further explore the use of a Large Language Model (LLM), i.e. GPT-4, for zero-shot re-ranking in various ToT domains, relying solely on the item titles. Results show that multi-domain training enhances recall, and that LLMs are strong zero-shot re-rankers, especially for popular items, outperforming direct GPT-4 prompting without first-stage retrieval. Datasets and code can be found on GitHub https://github.com/LuisPB7/TipTongue
- Open Access
Building Retrieval Systems for the ClueWeb22-B Corpus
The ClueWeb22 dataset containing nearly 10 billion documents was released in 2022 to support academic and industry research. The goal of this project was to build retrieval baselines for the English section of the "super head" part (category B) of this dataset. These baselines can then be used by the research community to compare their systems and also to generate data to train/evaluate new retrieval and ranking algorithms. The report covers sparse and dense first stage retrievals as well as neural rerankers that were implemented for this dataset. These systems are available as a service on a Carnegie Mellon University cluster.
Open AccessBuilding Retrieval Systems for the ClueWeb22-B Corpus
The ClueWeb22 dataset containing nearly 10 billion documents was released in 2022 to support academic and industry research. The goal of this project was to build retrieval baselines for the English section of the "super head" part (category B) of this dataset. These baselines can then be used by the research community to compare their systems and also to generate data to train/evaluate new retrieval and ranking algorithms. The report covers sparse and dense first stage retrievals as well as neural rerankers that were implemented for this dataset. These systems are available as a service on a Carnegie Mellon University cluster.
- Open Access
- 9 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 77-85
Conversational Search with Random Walks over Entity Graphs
The entities that emerge during a conversation can be used to model topics, but not all entities are equally useful for this task. Modeling the conversation with entity graphs and predicting each entity's centrality in the conversation provides additional information that improves the retrieval of answer passages for the current question. Experiments show that using random walks to estimate entity centrality on conversation entity graphs improves top precision answer passage ranking over competitive transformer-based baselines.
- 9 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 77-85
Open AccessConversational Search with Random Walks over Entity Graphs
The entities that emerge during a conversation can be used to model topics, but not all entities are equally useful for this task. Modeling the conversation with entity graphs and predicting each entity's centrality in the conversation provides additional information that improves the retrieval of answer passages for the current question. Experiments show that using random walks to estimate entity centrality on conversation entity graphs improves top precision answer passage ranking over competitive transformer-based baselines.
- Open Access
- 9 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 13-22
KALE: Using a K-Sparse Projector for Lexical Expansion
Recent research has proposed retrieval approaches based on sparse representations and inverted indexes, with terms produced by neural language models and leveraging the advantages from both neural retrieval and lexical matching. This paper proposes KALE, a new lightweight method of this family that uses a small model with a k-sparse projector to convert dense representations into a sparse set of entries from a latent vocabulary. The KALE vocabulary captures semantic concepts than perform well when used in isolation, and perform better when extending the original lexical vocabulary, this way improving first-stage retrieval accuracy. Experiments with the MSMARCOv1 passage retrieval dataset, the TREC Deep Learning dataset, and BEIR datasets, examined the effectiveness of KALE under varying conditions. Results show that the KALE terms can replace the original lexical vocabulary, with gains in accuracy and efficiency. Combining KALE with the original lexical vocabulary, or with other learned terms, can further improve retrieval accuracy with only a modest increase in computational cost.
- 9 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 13-22
Open AccessKALE: Using a K-Sparse Projector for Lexical Expansion
Recent research has proposed retrieval approaches based on sparse representations and inverted indexes, with terms produced by neural language models and leveraging the advantages from both neural retrieval and lexical matching. This paper proposes KALE, a new lightweight method of this family that uses a small model with a k-sparse projector to convert dense representations into a sparse set of entries from a latent vocabulary. The KALE vocabulary captures semantic concepts than perform well when used in isolation, and perform better when extending the original lexical vocabulary, this way improving first-stage retrieval accuracy. Experiments with the MSMARCOv1 passage retrieval dataset, the TREC Deep Learning dataset, and BEIR datasets, examined the effectiveness of KALE under varying conditions. Results show that the KALE terms can replace the original lexical vocabulary, with gains in accuracy and efficiency. Combining KALE with the original lexical vocabulary, or with other learned terms, can further improve retrieval accuracy with only a modest increase in computational cost.