Entity · paper

Improving Ad-hoc Search Effectiveness for Conversational Information Retrieval via Model Merging

paperactiveprovisionalimproving-ad-hoc-search-effectiveness-for-conversational-information-retrieval-via-model-merging-2f0c6d2f·1 events·first seen 8h ago

Aliases: Improving Ad-hoc Search Effectiveness for Conversational Information Retrieval via Model Merging

Co-occurring entities

Model Soup Slerp NDCG

More like this (12)

When Does Mixing Help? Analyzing Query Embedding Interpolation in Multilingual Dense Retrieval Self-Augmenting Retrieval for Diffusion Language Models Understanding the Behaviors of Environment-aware Information Retrieval Uncertainty-Aware Hybrid Retrieval for Long-Document RAG When Does Combining Language Models Help? A Co-Failure Ceiling on Routing, Voting, and Mixture-of-Agents Across 67 Frontier Models hybrid dense-sparse retrieval Contextual Retrieval Curated retrieval versus open web search in public AI information services: a coverage-trust trade-off Steerable Model Merging Recalling Too Well: Sycophancy Evaluation and Mitigation in Memory-Augmented Models CheckRLM: Effective Knowledge-Thought Coherence Checking in Retrieval-Augmented Reasoning Difference-Aware Retrieval Policies for Imitation Learning

Recent events (1)

4arXiv · cs.CL·8h ago·source ↗

Model merging as training-free strategy for conversational information retrieval

A new arXiv preprint proposes using model merging (Model Soup and Slerp) as a training-free alternative to fine-tuning for conversational information retrieval, avoiding catastrophic forgetting of ad-hoc retrieval capabilities. Experiments on standard ad-hoc and conversational retrieval benchmarks show up to 15% higher NDCG@3 under zero-shot conditions. The approach enables a single retrieval model to operate across both ad-hoc and conversational settings without retraining.

Evaluation and Benchmarking Model Soup Slerp Improving Ad-hoc Search Effectiveness for Conversational Information Retrieval via Model Merging +1 more