• Home
  • Categories
  • Tags
  • Pricing
  • Submit
    Decorative pattern
    1. Home
    2. Research Papers & Surveys
    3. LIR: Late Interaction Workshop @ ECIR 2026

    LIR: Late Interaction Workshop @ ECIR 2026

    The first workshop dedicated to late interaction and multi-vector retrieval methods at ECIR 2026, featuring keynote speaker Omar Khattab (ColBERT creator) and focusing on advances in token-level representations, multi-modal retrieval, and long-context search.

    🌐Visit Website

    About this tool

    Overview

    The LIR Workshop is the first academic workshop dedicated to late interaction and multi-vector retrieval methods, held at ECIR 2026 in Delft, Netherlands on April 2, 2026.

    Schedule

    • Date: April 2, 2026
    • Time: 09:00-13:00 (Half-day workshop)
    • Location: Delft, Netherlands (ECIR 2026 Workshop Day)

    Workshop Format

    • Keynote presentation
    • Paper sessions
    • Demo presentations
    • Roundtable discussion

    Keynote Speaker

    Omar Khattab, Assistant Professor at MIT, whose research pioneered modern multi-vector retrieval including:

    • ColBERT (original approach)
    • ColBERTv2 (improved version)
    • PLAID indexing method (de facto standard)

    Topics of Interest

    Core Topics

    • Late interaction retrieval methods
    • Multi-vector representations
    • Token-level similarity matching
    • Fine-grained document representations

    Advanced Research Areas

    1. Training and Optimization

      • Impact of different scheduling strategies
      • Data mix optimization for late-interaction models
      • Training loss functions
      • Model-specific improvements
    2. Multi-Modal Retrieval

      • ColPali for document understanding
      • VideoColBERT for video search
      • Modality-specific approaches
      • Cross-modal late interaction
    3. Long-Context and Agentic Tasks

      • Long-context retrieval challenges
      • Agentic Search systems
      • Deep Research tasks
      • Reasoning-based retrieval

    Why Late Interaction Matters

    Late interaction methods, pioneered by ColBERT, have emerged as a powerful alternative to single-vector neural IR by:

    • Leveraging fine-grained, token-level representations
    • Delivering strong generalization
    • Providing robust out-of-domain performance
    • Enabling explainable similarity matching

    Recent Advances

    • Early multimodal work suggests late-interaction models suit most modalities
    • ColPali and VideoColBERT showing promising results
    • Outperforming other retriever families in specific domains
    • Unified frameworks emerging for different modalities

    Community Impact

    The workshop addresses:

    • Fragmentation across research communities
    • Integration of practitioner perspectives
    • Rapid advances in multiple research areas
    • Need for standardized evaluation

    Call for Papers

    Welcoming submissions on:

    • Novel late-interaction architectures
    • Efficiency improvements
    • Multi-modal extensions
    • Real-world deployment experiences
    • Benchmark and evaluation studies

    Relevance to Industry

    Late interaction models are increasingly being adopted in production systems for:

    • Document retrieval without preprocessing
    • Visual question answering
    • Multi-modal search applications
    • Context-preserving similarity matching
    Surveys

    Loading more......

    Information

    Websitewww.lateinteraction.com
    PublishedMar 20, 2026

    Categories

    1 Item
    Research Papers & Surveys

    Tags

    3 Items
    #Workshop#Late Interaction#Academic

    Similar Products

    6 result(s)
    The Novel Vector Database

    Research paper proposing a decoupled storage architecture for vector databases that improves update speed by 10.05x for insertions and 6.89x for deletions through innovative design.

    When Large Language Models Meet Vector Databases: A Survey

    Comprehensive 2026 research survey examining the integration of LLMs with vector databases, addressing challenges like hallucination, bias, and real-time knowledge updates through vector-based retrieval.

    SPLATE

    Sparse Late Interaction Retrieval model that combines the benefits of sparse representations with late interaction mechanisms. Provides efficient storage and fast retrieval while maintaining the accuracy advantages of token-level matching in sparse embedding space.

    ColBERTv2
    Featured

    Advanced multi-vector retrieval model creating token-level embeddings with late interaction mechanism, featuring denoised supervision and improved memory efficiency over original ColBERT.

    ColQwen

    Late interaction retrieval model that applies the ColBERT token-level embedding approach using the Qwen language model as the base encoder. Provides high-quality semantic search with detailed token-level matching for improved retrieval accuracy.

    Nemotron ColEmbed V2

    State-of-the-art ColBERT-style embedding model family achieving top performance on ViDoRe benchmarks for visual document retrieval. The 8B model ranks first on ViDoRe V3 leaderboard with 63.42 average NDCG@10 as of February 2026.

    Decorative pattern
    Built with
    Ever Works
    Ever Works

    Connect with us

    Stay Updated

    Get the latest updates and exclusive content delivered to your inbox.

    Product

    • Categories
    • Tags
    • Pricing
    • Help

    Clients

    • Sign In
    • Register
    • Forgot password?

    Company

    • About Us
    • Admin
    • Sitemap

    Resources

    • Blog
    • Submit
    • API Documentation
    All product names, logos, and brands are the property of their respective owners. All company, product, and service names used in this repository, related repositories, and associated websites are for identification purposes only. The use of these names, logos, and brands does not imply endorsement, affiliation, or sponsorship. This directory may include content generated by artificial intelligence.
    Copyright © 2025 Awesome Vector Databases. All rights reserved.·Terms of Service·Privacy Policy·Cookies