Call for Shared Task Participation: AstroCLIMB (Astronomy Citation Linking from Illustrations, a Multimodal Benchmark)

--===============3972754291854343810==
Content-Type: multipart/alternative;
 boundary="=_04216a3c1d6257e7ba36d42ffa6c7f82"

--=_04216a3c1d6257e7ba36d42ffa6c7f82
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=US-ASCII;
 format=flowed

====================================================================

CALL FOR PARTICIPATION

AstroCLIMB: Astronomy Citation Linking from Illustrations, a Multimodal 
Benchmark

Shared task of the 4th Workshop on Artificial Intelligence for

Scientific Publications (WASP 2026), co-located with IJCNLP-AACL 2026

Workshop: November 9-10, 2026 (online)

Task page: https://ui.adsabs.harvard.edu/WIESP/2026/shared_task [1]

Kaggle competition: 
https://www.kaggle.com/competitions/astroclimb/overview [2]

Dataset: https://huggingface.co/datasets/adsabs/AstroCLIMB [3]

====================================================================

OVERVIEW

Scientists rely on figures to communicate their results, yet the 
information locked in those figures is hard to parse, archive, and 
search. Recent multimodal models promise to change that, but they have 
seen little testing on real scientific figures. AstroCLIMB, built in 
partnership with astroexplorer.org [4], offers a benchmark to probe 
exactly this capability: can a model reconstruct part of the citation 
graph of astronomy papers from their figures and captions alone?

The dataset comprises over 100K figure-and-caption pairs drawn from 
recent open-access astronomy papers.

TASK

Multi-class, single-label classification. Each input is a pair of 
objects: two figures, a figure and a caption, or two captions. The 
relation is symmetric, and each pair has exactly one label:

- same figure: the two objects come from the same scientific figure 
(figure-caption pairs only)

- same paper: the two objects come from the same paper (same DOI)

- related papers: one object's paper cites the other's

- unrelated papers: none of the above

A note on scope: AstroCLIMB is meant as a benchmark for multimodal 
models rather than an end in itself. Participants are not expected to 
train large models from scratch for it; computing basic statistics or 
folding it into a larger multimodal training pipeline is very much in 
keeping with the intent.

DATA

Two versions of the same dataset are available:

- Full dataset (Hugging Face): over 94K rows, each with a figure, its 
caption, and metadata (source DOI, title, authors, cited and citing 
DOIs) describing an astronomy citation graph as an adjacency list. Pairs 
are not enumerated; you generate them from the adjacency list.

   https://huggingface.co/datasets/adsabs/AstroCLIMB [3]

- Evaluation dataset (Kaggle): 10K enumerated pairs each in the train 
and test splits, with labels. Many more pairs can be generated from the 
full dataset.

   https://www.kaggle.com/competitions/astroclimb/overview [2]

In the Kaggle data, some objects are English-language strings, and 
others are images encoded as strings; a helper notebook for decoding 
them to images is provided on Kaggle.

EVALUATION

Systems are evaluated on Kaggle using macro-F1 over the four classes. 
Submissions are CSV files with per-class indicators; a sample submission 
is available on the competition page.

HOW TO PARTICIPATE

1. Register and join the competition on Kaggle to access the data and 
score submissions:

    https://www.kaggle.com/competitions/astroclimb/overview [2]

2. Optionally, write up your system description as a paper using the ACL 
LaTeX template

    (https://github.com/acl-org/acl-style-files [5]) and submit it via 
OpenReview:

    
https://openreview.net/group?id=aclweb.org/AACL-IJCNLP/2026/Workshop/WASP 
[6]

System description papers undergo light peer review and, if accepted, 
are published in the WASP 2026 proceedings in the ACL Anthology. We 
encourage a submission regardless of leaderboard position.

TIMELINE (subject to change; all deadlines 11.59 pm UTC-12h, anywhere on 
Earth)

- 1st CfP + Registration opens: July 27, 2026

- Train and scoring portion of test data release: July 27, 2026

- Registration ends: September 12, 2026

- System run and output submission: September 13, 2026

- System paper submission: September 14, 2026

- Result announcement: October 1, 2026

- Camera-ready submission: October 12, 2026

- Full data release: November 2, 2026

- Workshop: November 9-10, 2026

CONTACT

For inquiries, contact Felix Grezes at [email protected]

We look forward to your participation.

Links:
------
[1] https://ui.adsabs.harvard.edu/WIESP/2026/shared_task
[2] https://www.kaggle.com/competitions/astroclimb/overview
[3] https://huggingface.co/datasets/adsabs/AstroCLIMB
[4] http://astroexplorer.org/
[5] https://github.com/acl-org/acl-style-files
[6] 
https://openreview.net/group?id=aclweb.org/AACL-IJCNLP/2026/Workshop/WASP
--=_04216a3c1d6257e7ba36d42ffa6c7f82
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html; charset=UTF-8

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html; charset=
=3DUTF-8" /></head><body style=3D'font-size: 10pt; font-family: Verdana,Gen=
eva,sans-serif'>
<p id=3D"m_-2556216839184460982gmail-docs-internal-guid-33f7024a-7fff-429d-=
2ec5-0ca9b2bc8f99" dir=3D"ltr" style=3D"color: #222222; font-size: small; f=
ont-family: Arial, Helvetica, sans-serif; background-color: #ffffff;"><span=
 style=3D"font-size: 11pt; font-family: Arial, sans-serif; background-color=
: transparent;">=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D<wbr />=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D<wbr />=3D=3D=3D=
=3D=3D=3D=3D=3D</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; font-weight: bold; background-co=
lor: transparent;">CALL FOR PARTICIPATION</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
AstroCLIMB: Astronomy Citation Linking from Illustrations, </span><span sty=
le=3D"font-size: 11pt; font-family: Arial, sans-serif; background-color: tr=
ansparent;">a Multimodal Benchmark</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Shared task of the 4th Workshop on Artificial Intelligence for</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Scientific Publications (WASP 2026), co-located with IJCNLP-AACL 2026</span=
></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Workshop: November 9-10, 2026 (online)</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Task page: </span><a style=3D"color: #1155cc;" href=3D"https://ui.adsabs.ha=
rvard.edu/WIESP/2026/shared_task" target=3D"_blank" rel=3D"noopener" data-s=
aferedirecturl=3D"https://www.google.com/url?q=3Dhttps://ui.adsabs.harvard.=
edu/WIESP/2026/shared_task&amp;source=3Dgmail&amp;ust=3D1785279469448000&am=
p;usg=3DAOvVaw1RhTMpntloPTt_ZRdN1SQU"><span style=3D"color: #1155cc; font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
https://ui.adsabs.harvard.edu/<wbr />WIESP/2026/shared_task</span></a></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Kaggle competition: </span><a style=3D"color: #1155cc;" href=3D"https://www=
=2Ekaggle.com/competitions/astroclimb/overview" target=3D"_blank" rel=3D"no=
opener" data-saferedirecturl=3D"https://www.google.com/url?q=3Dhttps://www.=
kaggle.com/competitions/astroclimb/overview&amp;source=3Dgmail&amp;ust=3D17=
85279469448000&amp;usg=3DAOvVaw0VYkNiRiZTIyl71i9dJGDz"><span style=3D"color=
: #1155cc; font-size: 11pt; font-family: Arial, sans-serif; background-colo=
r: transparent;">https://www.kaggle.com/<wbr />competitions/astroclimb/<wbr=
 />overview</span></a></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Dataset: </span><a style=3D"color: #1155cc;" href=3D"https://huggingface.co=
/datasets/adsabs/AstroCLIMB" target=3D"_blank" rel=3D"noopener" data-safere=
directurl=3D"https://www.google.com/url?q=3Dhttps://huggingface.co/datasets=
/adsabs/AstroCLIMB&amp;source=3Dgmail&amp;ust=3D1785279469448000&amp;usg=3D=
AOvVaw1qPrh3Z5Z_vJEBoDrihWy5"><span style=3D"color: #1155cc; font-size: 11p=
t; font-family: Arial, sans-serif; background-color: transparent;">https://=
huggingface.co/<wbr />datasets/adsabs/AstroCLIMB</span></a></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D<wbr />=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D<wbr />=3D=3D=3D=3D=3D=3D=3D=3D</spa=
n></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; font-weight: bold; background-co=
lor: transparent;">OVERVIEW</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Scientists rely on figures to communicate their results, yet the informatio=
n locked in those figures is hard to parse, archive, and search. Recent mul=
timodal models promise to change that, but they have seen little testing on=
 real scientific figures. AstroCLIMB, built in partnership with <span aria-=
invalid=3D"grammar"><a style=3D"color: #1155cc;" href=3D"http://astroexplor=
er.org/" target=3D"_blank" rel=3D"noopener" data-saferedirecturl=3D"https:/=
/www.google.com/url?q=3Dhttp://astroexplorer.org&amp;source=3Dgmail&amp;ust=
=3D1785279469448000&amp;usg=3DAOvVaw1p0XagrE0MJ-YV7JwO_J_I">astroexplorer.o=
rg</a></span>, offers a benchmark to probe exactly this capability: can a m=
odel reconstruct part of the citation graph of astronomy papers from their =
figures and captions alone?</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
The dataset comprises over 100K figure-and-caption pairs drawn from recent =
open-access astronomy papers.</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; font-weight: bold; background-co=
lor: transparent;">TASK</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Multi-class, single-label classification. Each input is a pair of objects: =
two figures, a figure and a caption, or two captions. The relation is symme=
tric, and each pair has exactly one label:</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- same figure: the two objects come from the same scientific figure (figure=
-caption pairs only)</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- same paper: the two objects come from the same paper (same DOI)</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- related papers: one object's paper cites the other's</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- unrelated papers: none of the above</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
A note on scope: AstroCLIMB is meant as a benchmark for multimodal models r=
ather than an end in itself. Participants are not expected to train large m=
odels from scratch for it; computing basic statistics or folding it into a =
larger multimodal training pipeline is very much in keeping with the intent=
=2E</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; font-weight: bold; background-co=
lor: transparent;">DATA</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Two versions of the same dataset are available:</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- Full dataset (Hugging Face): over 94K rows, each with a figure, its capti=
on, and metadata (source DOI, title, authors, cited and citing DOIs) descri=
bing an astronomy citation graph as an adjacency list. Pairs are not enumer=
ated; you generate them from the adjacency list.</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
&nbsp;&nbsp;</span><a style=3D"color: #1155cc;" href=3D"https://huggingface=
=2Eco/datasets/adsabs/AstroCLIMB" target=3D"_blank" rel=3D"noopener" data-s=
aferedirecturl=3D"https://www.google.com/url?q=3Dhttps://huggingface.co/dat=
asets/adsabs/AstroCLIMB&amp;source=3Dgmail&amp;ust=3D1785279469448000&amp;u=
sg=3DAOvVaw1qPrh3Z5Z_vJEBoDrihWy5"><span style=3D"color: #1155cc; font-size=
: 11pt; font-family: Arial, sans-serif; background-color: transparent;">htt=
ps://huggingface.co/<wbr />datasets/adsabs/AstroCLIMB</span></a></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- Evaluation dataset (Kaggle): 10K enumerated pairs each in the train and t=
est splits, with labels. Many more pairs can be generated from the full dat=
aset.</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
&nbsp;&nbsp;</span><a style=3D"color: #1155cc;" href=3D"https://www.kaggle.=
com/competitions/astroclimb/overview" target=3D"_blank" rel=3D"noopener" da=
ta-saferedirecturl=3D"https://www.google.com/url?q=3Dhttps://www.kaggle.com=
/competitions/astroclimb/overview&amp;source=3Dgmail&amp;ust=3D178527946944=
8000&amp;usg=3DAOvVaw0VYkNiRiZTIyl71i9dJGDz"><span style=3D"color: #1155cc;=
 font-size: 11pt; font-family: Arial, sans-serif; background-color: transpa=
rent;">https://www.kaggle.com/<wbr />competitions/astroclimb/<wbr />overvie=
w</span></a></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
In the Kaggle data, some objects are English-language strings, and others a=
re images encoded as strings; a helper notebook for decoding them to images=
 is provided on Kaggle.</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; font-weight: bold; background-co=
lor: transparent;">EVALUATION</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
Systems are evaluated on Kaggle using macro-F1 over the four classes. Submi=
ssions are CSV files with per-class indicators; a sample submission is avai=
lable on the competition page.</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; font-weight: bold; background-co=
lor: transparent;">HOW TO PARTICIPATE</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
1. Register and join the competition on Kaggle to access the data and score=
 submissions:</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
&nbsp;&nbsp;&nbsp;</span><a style=3D"color: #1155cc;" href=3D"https://www.k=
aggle.com/competitions/astroclimb/overview" target=3D"_blank" rel=3D"noopen=
er" data-saferedirecturl=3D"https://www.google.com/url?q=3Dhttps://www.kagg=
le.com/competitions/astroclimb/overview&amp;source=3Dgmail&amp;ust=3D178527=
9469448000&amp;usg=3DAOvVaw0VYkNiRiZTIyl71i9dJGDz"><span style=3D"color: #1=
155cc; font-size: 11pt; font-family: Arial, sans-serif; background-color: t=
ransparent;">https://www.kaggle.com/<wbr />competitions/astroclimb/<wbr />o=
verview</span></a></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
2. Optionally, write up your system description as a paper using the ACL La=
TeX template</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
&nbsp;&nbsp;&nbsp;(</span><a style=3D"color: #1155cc;" href=3D"https://gith=
ub.com/acl-org/acl-style-files" target=3D"_blank" rel=3D"noopener" data-saf=
eredirecturl=3D"https://www.google.com/url?q=3Dhttps://github.com/acl-org/a=
cl-style-files&amp;source=3Dgmail&amp;ust=3D1785279469448000&amp;usg=3DAOvV=
aw1jkO2Cb-AFDvm79MMp8vDf"><span style=3D"color: #1155cc; font-size: 11pt; f=
ont-family: Arial, sans-serif; background-color: transparent;">https://gith=
ub.com/acl-<wbr />org/acl-style-files</span></a><span style=3D"font-size: 1=
1pt; font-family: Arial, sans-serif; background-color: transparent;">) and =
submit it via OpenReview:</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
&nbsp;&nbsp;&nbsp;</span><a style=3D"color: #1155cc;" href=3D"https://openr=
eview.net/group?id=3Daclweb.org/AACL-IJCNLP/2026/Workshop/WASP" target=3D"_=
blank" rel=3D"noopener" data-saferedirecturl=3D"https://www.google.com/url?=
q=3Dhttps://openreview.net/group?id%3Daclweb.org/AACL-IJCNLP/2026/Workshop/=
WASP&amp;source=3Dgmail&amp;ust=3D1785279469448000&amp;usg=3DAOvVaw3EMlGNIq=
-QMJs-d--0qBCk"><span style=3D"color: #1155cc; font-size: 11pt; font-family=
: Arial, sans-serif; background-color: transparent;">https://openreview.net=
/<wbr />group?id=3Daclweb.org/AACL-<wbr />IJCNLP/2026/Workshop/WASP</span><=
/a></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
System description papers undergo light peer review and, if accepted, are p=
ublished in the WASP 2026 proceedings in the ACL Anthology. We <span aria-i=
nvalid=3D"grammar">encourage a submission</span> regardless of leaderboard =
position.</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
<strong>TIMELINE</strong> (subject to change; all deadlines 11.59 pm UTC-12=
h, anywhere on Earth)</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- 1st CfP + Registration opens: July 27, 2026</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- Train and scoring portion of test data release: July 27, 2026</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- Registration ends: September 12, 2026</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- System run and output submission: September 13, 2026</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- System paper submission: September 14, 2026</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- Result announcement: October 1, 2026</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- Camera-ready submission: October 12, 2026</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- Full data release: November 2, 2026</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
- Workshop: November 9-10, 2026</span></p>
<p><br /></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; font-weight: bold; background-co=
lor: transparent;">CONTACT</span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
For inquiries, contact Felix Grezes at </span><span style=3D"font-size: 11p=
t; font-family: Arial, sans-serif; font-weight: bold; background-color: tra=
nsparent;"><a style=3D"color: #1155cc;" href=3D"mailto:[email protected]=
vard.edu" target=3D"_blank" rel=3D"noopener">[email protected]</=
a></span></p>
<p dir=3D"ltr" style=3D"color: #222222; font-size: small; font-family: Aria=
l, Helvetica, sans-serif; background-color: #ffffff;"><span style=3D"font-s=
ize: 11pt; font-family: Arial, sans-serif; background-color: transparent;">=
We look forward to your participation.</span></p>

</body></html>

--=_04216a3c1d6257e7ba36d42ffa6c7f82--

--===============3972754291854343810==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Corpora mailing list -- [email protected]
https://list.elra.info/mailman3/postorius/lists/corpora.list.elra.info/
To unsubscribe send an email to [email protected]

--===============3972754291854343810==--