Overview

At a Glance

Why submit?

  • Reach a multidisciplinary audience spanning NLP, ML, and IR communities.
  • Choose between direct submission and an ARR commitment fast track.
  • Publish archival proceedings papers or present non-archival work for community discussion.
  • Showcase new benchmarks, datasets, systems, and lessons learned — including negative results.

Who should submit?

  • Students, researchers, and practitioners in NLP, IR, and ML.
  • Work on retrieval-augmented generation, knowledge grounding, factuality, and hallucination.
  • Efficient training, adaptation, and inference methods for grounded LLMs.
  • Evaluation frameworks and benchmarks for grounding and faithfulness.

Workshop format

  • Full-day workshop with invited keynote talks.
  • Panel discussion on open challenges in grounding.
  • Contributed oral presentations and poster session.
  • Student mentoring and networking session.
Scope

Topics of Interest

Submissions

Call for Papers

We invite submissions of original research on all aspects of grounding in language models, with a particular emphasis on faithfulness, efficiency, and real-world reliability. To provide maximum flexibility, GroundLM 2026 offers two submission tracks: direct submission for new work and an ARR commitment fast track.

Track 1

Direct Submission

Deadline: June 29, 2026 (AoE)

For new work submitted directly to GroundLM through OpenReview. Authors may submit archival or non-archival papers.

Track 2

ARR Commitment

Deadline: August 5, 2026 (AoE)

A fast track for papers from the ARR May 2026 cycle or earlier, as well as papers rejected from other conferences.

  • For ARR papers, we recommend an average Overall Assessment score of at least 2.5, or a meta-review score of at least 2.5.
  • Papers rejected from other conferences should include the reviews and meta-reviews, or AC/SAC decisions, with the submission.
  • Authors should commit their reviews and meta-reviews through OpenReview. The senior program committee will make rapid final decisions based on these reviews.

Submission Types

  • Long papers (archival): Up to 8 pages of content, unlimited references. Accepted papers will appear in ACL Anthology proceedings.
  • Short papers (archival): Up to 4 pages of content, unlimited references. Accepted papers will appear in ACL Anthology proceedings.
  • Non-archival papers: No page limit, though 4-8 pages is standard. These papers may be presented at the workshop but will not appear in the proceedings.
  • All accepted archival and non-archival papers are eligible for oral or poster presentation.

Submission Instructions

  • All submissions must follow the ACL 2026 style guidelines and be anonymized for double-blind review.
  • Submit via OpenReview.
  • Interested reviewers and area chairs may self-nominate through the reviewer and area chair self-nomination form.
  • Reviewing is double-blind. Author names and affiliations must be omitted.
  • Track 1 archival submissions must not be under review at other venues for the duration of the GroundLM review period.
  • We follow the ACL Code of Ethics. Authors are expected to attest to the ethical considerations in their work.

Dual Submission & Preprints

Non-archival submissions may overlap with previously published or concurrently submitted work. Preprints may be posted publicly at any time without violating the anonymity policy.

Best Paper Award

Outstanding submissions will be recognized with a Best Paper Award, selected by the program committee. We particularly encourage strong empirical work, novel theoretical contributions, and resource papers.

Track 1 Submission → Track 2 ARR Commitment → Reviewer / AC Self-Nomination →
Timeline

Important Dates for Paper Submissions

Paper submission deadlines are end-of-day in Anywhere on Earth (AoE) time.

EventDate

Shared task dates are listed separately on the Shared Tasks page.

Challenges

Shared Tasks

GroundLM 2026 hosts two shared tasks for evaluating faithful grounding across visual scenes and scientific literature. Public development sets and local evaluators are available.

Shared Task 1

GoldenViewVQA

Multi-view visual question answering with evidence-source identification in driving scenes.

Shared Task 2

LitTraceQA

Literature-grounded question answering with paper retrieval, evidence grounding, and answer generation.

View Shared Tasks
Keynotes

Invited Speakers

More speakers to be announced. Check back soon.

Team

Organizers

Questions? Contact us at groundlm2026@googlegroups.com

Schedule

Program

The detailed program will be available after acceptance notifications (mid August 2026). The schedule below is a tentative outline.

Workshop Day — October 29, 2026, 9:00-17:30

Location

Venue

EMNLP 2026

GroundLM 2026 is co-located with EMNLP 2026. The room and exact location will be announced by the EMNLP 2026 organizing committee.

EMNLP (Empirical Methods in Natural Language Processing) is one of the premier annual conferences in NLP. GroundLM 2026 will take place on October 29, 2026, from 9:00 to 17:30. Details to follow.

For travel and accommodation information, please refer to the EMNLP 2026 website once it becomes available.

Accessibility

We are committed to making GroundLM 2026 accessible and inclusive. Please contact us if you have specific accessibility needs.

Venue TBD — EMNLP 2026 October 29, 2026 · 9:00-17:30
Questions

FAQ