Runs - Interactive Knowledge Assistance Track (IKAT) 2025¶

agg_false-qrec-mse-sum-top10¶

Run ID: agg_false-qrec-mse-sum-top10
Participant: guidance
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: 65d0586c6ae14d4c411f07568ce619ba

agg_true-qrec-mse-sum-top10¶

Run ID: agg_true-qrec-mse-sum-top10
Participant: guidance
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: 533fec4381f1d36ec37408f024dd47a4

cfda-adarewriter-chiq-llm4cs-splade¶

Run ID: cfda-adarewriter-chiq-llm4cs-splade
Participant: cfda
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: a344ef36b979111671b762c5519d4c38
Run description: Rewrites queries with AdaRewriter (CHIQ-AD + LLM4CS), retrieves passages via SPLADE, reranks top-2000 with DeBERTaV3 cross-encoder, then generates answers with GPT4o.

cfda-auto-1¶

Run ID: cfda-auto-1
Participant: cfda
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: 7fe3d2fa637d454af7566630b7feb0e3

cfda-auto-2¶

Run ID: cfda-auto-2
Participant: cfda
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: ffdcb34baf63d55fb2fff47a28276d7e

cfda-auto-3¶

Run ID: cfda-auto-3
Participant: cfda
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: 2caf0f92e09a3e43c5c63a515c6be9eb

cfda-auto-4¶

Run ID: cfda-auto-4
Participant: cfda
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: 5ccb4987a5b134e020b133c878fa1347

cfda-chiq-llm4cs-splade-rrf¶

Run ID: cfda-chiq-llm4cs-splade-rrf
Participant: cfda
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: 376857d360e374f541a9702c6f2b5e38
Run description: Generates 2 rewrites each with CHIQ-AD and LLM4CS, retrieves passages with SPLADE, fuses results using RRF, then generates response with GPT-4o.

cfda-gen-only-1¶

Participants | Proceedings | Input | Appendix

Run ID: cfda-gen-only-1
Participant: cfda
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-gen
MD5: c4f6a5c9613a04001eb412d00b95f901

cfda-gen-only-2¶

Participants | Proceedings | Input | Appendix

Run ID: cfda-gen-only-2
Participant: cfda
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-gen
MD5: fdd927ee0c7d90f4fe422de1cd93334c

cosine-orconvqa-sum-top10¶

Run ID: cosine-orconvqa-sum-top10
Participant: guidance
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: 93106890394fc889cb71bd53f914d4db

disco-qrecc-norerank¶

Run ID: disco-qrecc-norerank
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: cbc46a1a80257144efd95ece9176ee33

genaius-full-rewrite¶

Run ID: genaius-full-rewrite
Participant: GenAIus
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: bd2366d1f6bfab2cf789308dbc385761
Run description: FullHistory-RewriteUtterance-BM25-GPT4o

genaius-genonly-full-gpt4o¶

Participants | Proceedings | Input | Appendix

Run ID: genaius-genonly-full-gpt4o
Participant: GenAIus
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-08-11
Task: trec2025-ikat-gen
MD5: 27294f6261fee4ee16986a1c30a669b0

genaius-genonly-summary-gpt4o¶

Participants | Proceedings | Input | Appendix

Run ID: genaius-genonly-summary-gpt4o
Participant: GenAIus
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-08-11
Task: trec2025-ikat-gen
MD5: 7792a03d36590932af832ee1bf4f903e

genaius-summary-rewrite¶

Run ID: genaius-summary-rewrite
Participant: GenAIus
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: e64873d37e471ac648d7e83bb6c9ed70
Run description: SummaryHistory-RewriteUtterance-BM25-GPT4o

genonly-noptkb¶

Participants | Proceedings | Input | Appendix

Run ID: genonly-noptkb
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-gen
MD5: 161835351a7bd49352d9d85f9d4d2a21

genonly-ptkb¶

Participants | Proceedings | Input | Appendix

Run ID: genonly-ptkb
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-gen
MD5: aa586ada4afe5017a0a6e8ed5383a9d2

gpt-clarif-sum-top10¶

Participants | Proceedings | Input | Appendix

Run ID: gpt-clarif-sum-top10
Participant: guidance
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-08-20
Task: trec2025-ikat-gen
MD5: eef57a0f14dea073515944dbee82866a

grilllab-agentic-gpt4.1¶

Run ID: grilllab-agentic-gpt4.1
Participant: grilllab
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: 1e094a6d0d08e0f37925399154a2d510
Run description: Agentic pipeline where an LLM decides an action to take based on the conversation context, manual orchestration is minimal. Agent has access to search engine as a tool.

grilllab-agentic-gpt4.1-larf¶

Run ID: grilllab-agentic-gpt4.1-larf
Participant: grilllab
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: a5f541b4682f5e26fdb4859afb2a752b
Run description: Agentic pipeline where an LLM decides an action to take based on the conversation context, manual orchestration is minimal. Agent has access to a LARF based search engine as a tool.

grilllab-agentic-gpt4.1-larf-v2¶

Run ID: grilllab-agentic-gpt4.1-larf-v2
Participant: grilllab
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: cfc76ab528016f3302a06db8da16f8fc
Run description: Agentic pipeline where an LLM decides an action to take based on the conversation context, manual orchestration is minimal. Agent has access to a LARF based search engine as a tool.

grilllab-larf-fine-tuned-judge¶

Run ID: grilllab-larf-fine-tuned-judge
Participant: grilllab
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: cb856e45a5dcf0e3e1bc01550edf42eb
Run description: Three stage pipeline where we retrieve a candidate pool, expand the candidate pool with doc2query techniques, and refine the pool with an LLM. Each stage has a sub-step where we assess the candidate pool with respect to the query and filter out passages that are not relevant, followed by a reranking step.

grilllab-larf-finetuned¶

Run ID: grilllab-larf-finetuned
Participant: grilllab
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-26
Task: trec2025-ikat-auto
MD5: fab1d2541e65507f297c4e34d16f0799

grilllab-larf-finetuned-10-rounds¶

Run ID: grilllab-larf-finetuned-10-rounds
Participant: grilllab
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: 5db3ab94ffcb25fafe8a71c7022dbd61

grilllab-larf-finetuned-22-rounds¶

Run ID: grilllab-larf-finetuned-22-rounds
Participant: grilllab
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-08-01
Task: trec2025-ikat-auto
MD5: 8774724a2dd8b0e6a249fd3d9b985eb4

grilllab-larf-finetuned-rankllm¶

Run ID: grilllab-larf-finetuned-rankllm
Participant: grilllab
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: 4c76d63d95ef4063806d9a7acbe126bf

mq4cs-gpt41-bm25¶

Run ID: mq4cs-gpt41-bm25
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: 3c705c684207c567b853e7433740e871

mq4cs-gpt41-splade¶

Run ID: mq4cs-gpt41-splade
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: e593c3dc9f7123d02a30e54ad1d7ec28

mq4cs-llamaft-splade¶

Run ID: mq4cs-llamaft-splade
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-27
Task: trec2025-ikat-auto
MD5: 366953dffb654345f6c9b464322f600e

orga-gpt41mini-bm25-minilm-llama70b¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: orga-gpt41mini-bm25-minilm-llama70b
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: c9be418667ac8f32f966a3bedb6333be
Run description: The approach uses a RAG pipeline with Llama 70B as generation backbone, with gpt-4.1-mini for the rewrite, followed by bm25 and miniLM.

orga-gpt41mini-bm25-minilm-llama70b-nopersonal¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: orga-gpt41mini-bm25-minilm-llama70b-nopersonal
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: 520c4b24fe349d7b31b305a5413d44e5
Run description: This approach is not usign the PTKB. The approach uses a RAG pipeline with Llama 70B as generation backbone, with gpt-4.1-mini for the rewrite, followed by bm25 and miniLM.

orga-llama70b-bm25-minilm-llama70b¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: orga-llama70b-bm25-minilm-llama70b
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: e605353a848481ab72ed33abc82da26e
Run description: The approach uses a RAG pipeline with llama3.3 70B as generation backbone, with Llama 70B for the rewrite, followed by bm25 and miniLM.

orga-llama8b-bm25-minilm-llama8b-v2¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: orga-llama8b-bm25-minilm-llama8b-v2
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: 8e5511a8f5e1ed448ed7bc0f257207e7
Run description: The approach uses a RAG pipeline with llama3 8B as generation backbone, with Llama 8B for the rewrite, followed by bm25 and miniLM.

orga-no-no-no-gpt41mini¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: orga-no-no-no-gpt41mini
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: 6bf7cb77ac157606510dc73f2469ca3f
Run description: The approach uses gpt-4.1-mini as generation backbone, without retrieval or rewriting.

orga-no-no-no-llama70b¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: orga-no-no-no-llama70b
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: ab9b58b7d22f83cef0195c6992a8f4b5
Run description: The approach uses Llama 3.3 70B as generation backbone, without retrieval or rewriting.

orga-no-no-no-llama8b-v2¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: orga-no-no-no-llama8b-v2
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: 1d83bd02189404bb267c2ec1dd8b9d12
Run description: The approach uses Llama3 8B as generation backbone, without retrieval or rewriting.

organizers-baseline-gpt41mini-bm25-minilm-llama70b-gpt41mini¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: organizers-baseline-gpt41mini-bm25-minilm-llama70b-gpt41mini
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: 133734880aaf4ca42ca607e44df14d08

organizers-baseline-gpt41mini-bm25-minilm-llama70b-gpt41mini_nopersonal¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: organizers-baseline-gpt41mini-bm25-minilm-llama70b-gpt41mini_nopersonal
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: 92bd6916339c7ac19e2b45784bed8d71

organizers-baseline-gpto4mini-ance-norerank-llama70b-gpt41mini¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: organizers-baseline-gpto4mini-ance-norerank-llama70b-gpt41mini
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: ad05d65a3d825e7dd6aa5860cdfb2ae9

organizers-baseline-gpto4mini-splade-norerank-llama70b-gpt41mini¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: organizers-baseline-gpto4mini-splade-norerank-llama70b-gpt41mini
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: 96f15ee4d06f9c254a09a1d9b35a7d9b

organizers-baseline-llama70B-ance-minilm-llama70b-gpt41mini¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: organizers-baseline-llama70B-ance-minilm-llama70b-gpt41mini
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: fde833b38786a94a35f5938991665a7e

organizers-baseline-llama70B-splade-minilm-llama70b-gpt41mini¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: organizers-baseline-llama70B-splade-minilm-llama70b-gpt41mini
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: daeb061a471d9ddf00d0645377300660

organizers-baseline-manual-bm25-minilm-llama70b-gpt41mini¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: organizers-baseline-manual-bm25-minilm-llama70b-gpt41mini
Participant: coordinators
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: d8eb3b664ff531dbc095050f397ab509

ucsc-base-dynamicPTKB-trainedReranker¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: ucsc-base-dynamicPTKB-trainedReranker
Participant: ucsc
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: 41a772d9a3e896365521aeb3c91c0858
Run description: This approach utilizes LLMs to extract additional relevant PTKB statements from the conversation. Query generation is done using the base query, and the reranking is based on a model trained on previous TREC data.

UCSC-base-ensemble¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: UCSC-base-ensemble
Participant: ucsc
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: b5eb15ab46c76c28ad99db1b4e9cd733

UCSC-base-trained-MiniLM¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: UCSC-base-trained-MiniLM
Participant: ucsc
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: b6f42532fb382792b642908e255b636a

ucsc-SIMRAG-guidelineQuery-dynamicPTKB-trainedReranker¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: ucsc-SIMRAG-guidelineQuery-dynamicPTKB-trainedReranker
Participant: ucsc
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: fec055c382ce56b5adba6ca240e5b53a
Run description: This approach utilizes LLMs to extract additional relevant PTKB statements from the conversation. It then utilizes a trained SIMRAG model to obtain a good guideline-generated retrieval query, along with a good reranking query. The reranking is based on a model trained on previous TREC data.

UCSC-SIMRAG-keyword-ensemble¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: UCSC-SIMRAG-keyword-ensemble
Participant: ucsc
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: 15e5e21e25d043f3998ab76c13a3cccc

UCSC-SIMRAG-keyword-trained-MiniLM¶

Participants | Input | ptkb_trec_eval | docs_trec_eval | Appendix

Run ID: UCSC-SIMRAG-keyword-trained-MiniLM
Participant: ucsc
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-28
Task: trec2025-ikat-auto
MD5: aed20ff2358d5ce9a9bdc9fd26f45c22

ucsc-SIMRAG-keywordQuery-dynamicPTKB-ensembleReranker¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: ucsc-SIMRAG-keywordQuery-dynamicPTKB-ensembleReranker
Participant: ucsc
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: d8602007b2b8a199e998013a1eb1c964
Run description: This approach utilizes LLMs to extract additional relevant PTKB statements from the conversation. It then utilizes a trained SIMRAG model to obtain a good keyword retrieval query, along with a good reranking query. The reranking is based on an ensemble of five rerankers, reranking the top 100.

ucsc-SIMRAG-keywordQuery-dynamicPTKB-trainedReranker¶

Participants | Input | human_eval | retrieval_eval | Appendix

Run ID: ucsc-SIMRAG-keywordQuery-dynamicPTKB-trainedReranker
Participant: ucsc
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: b9da76dd67108ad88399283d2dc3ebba
Run description: This approach utilizes LLMs to extract additional relevant PTKB statements from the conversation. It then utilizes a trained SIMRAG model to obtain a good keyword retrieval query, along with a good reranking query. The reranking is based on a model trained on previous TREC data.

usiir_run1¶

Participants | Proceedings | Input | Appendix

Run ID: usiir_run1
Participant: usiir
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-29
Task: trec2025-ikat-gen
MD5: 14806efe08cb50a3f7556173bd3c8788

usiir_run2¶

Participants | Proceedings | Input | Appendix

Run ID: usiir_run2
Participant: usiir
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-07-29
Task: trec2025-ikat-gen
MD5: c4e2722b182487d125f2dd71320fd5f9

uva-gpt5-bm25-debertav3-gpt5¶

Run ID: uva-gpt5-bm25-debertav3-gpt5
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: 2d6ebf7d0843dd6bc00f5fd35313b337
Run description: The approach uses a RAG pipeline with gpt 5 as generation backbone, with gpt 5 for the rewrite, followed by bm25 and debertav3.

uva-gpt5-bm25-debertav3-gpt5mini-nopersonal¶

Run ID: uva-gpt5-bm25-debertav3-gpt5mini-nopersonal
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: b091860d0983e477ef8735bac48783df
Run description: The approach uses a RAG pipeline with gpt 5 mini as generation backbone, with gpt 5 mini for the rewrite, followed by bm25 and debertav3. This one doesn't use ptkb.

uva-gpt5mini-bm25-debertav3-gpt5mini¶

Run ID: uva-gpt5mini-bm25-debertav3-gpt5mini
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: 8aede87a1ab6fd6607a4670c722d9519
Run description: The approach uses a RAG pipeline with gpt 5 mini as generation backbone, with gpt 5 mini for the rewrite, followed by bm25 and debertav3.

uva-gpt5mini-no-no-gpt5mini¶

Run ID: uva-gpt5mini-no-no-gpt5mini
Participant: uva
Track: Interactive Knowledge Assistance Track (IKAT)
Year: 2025
Submission: 2025-12-01
Type: interactive
Task: trec2025-ikat-sim
MD5: ce7373e3891db400af40070fac0a4ef0
Run description: The approach uses a closed-book pipeline with gpt 5 mini as generation backbone.