Tanvi Dinkar
@t_dinkar
Post doc in Ethics in Conversational AI @HeriotWattUni | Background in CS and Linguistics. Robustness, Ethics, Dialogue, Pragmatics
Come check out our poster on anthropomorphism in dialogue systems @emnlpmeeting on the 9th at 11.00! @gavin_does_nlp @CurriedAmanda @ZeerakTalat @verena_rieser Also thank you to the reviewers and @emilymbender for their insightful feedback! Paper: arxiv.org/pdf/2305.09800…

7) Re-examining Sexism and Misogyny Classification with Annotator Attitudes (EMNLP 2024 findings) @AggieJiang @NikVits @t_dinkar @gavin_does_nlp @sinantie
Consistency is Key: Disentangling Label Variation in Natural Language Processing with Intra-Annotator Agreement arxiv.org/abs/2301.10684 @gavin_does_nlp, @t_dinkar, @CurriedAmanda, @verena_rieser and @dirk_hovy (6/6)
Thrilled that our work is mentioned! We found that annotators struggled to understand what was meant by 'fluency', and in turn it affected the reproducibility of the annotations. @gavin_does_nlp @verena_rieser
New blog: Human eval: Subjects must understand the task ehudreiter.com/2024/05/28/hum… This sounds obvious, but it was repeatedly raised as a concern in the replication shared task at the COLING human evaluation workshop.
Well done Sachin for publishing a paper based on his masters thesis in #HumEval2024 👏👏. The work tried to reproduce annotations in code-mixed Malayalam-English data! @heysachins @gavin_does_nlp @LrecColing
"Exploring Reproducibility of Human-Labelled Data for Code-Mixed Sentiment Analysis". Great virtual presentation by @heysachins. An interesting reproduction of both a training corpus and the classifiers which are then trained on it.
We go back to December (all the time) with @t_dinkar @gavin_does_nlp @ZeerakTalat and @verena_rieser (extra points for the @taylorswift13 reference)
#ThrowbackThursday #NLProc "Mirages. On Anthropomorphism in Dialogue Systems" by Abercrombie et al. examines linguistic factors contributing to anthropomorphism in conversational systems, with recommendations for responsible design. aclanthology.org/2023.emnlp-mai…
Now we have our first ReproNLP2024 paired presentation, @t_dinkar and Javier González Corbelle presenting their reproduction studies of aclanthology.org/2021.acl-long. #HumEval2024.
Excited for this year’s speakers! If you are attending LREC-COLING in a couple of weeks, join our workshop!
📢 Keynotes for the Safety in Conversational AI workshop @LrecColing! 📢 Upcoming talks from Laura Weidinger, Stevie Bergman, and Maurice Jakesch! 🎉🎉 Stay tuned for details 👏 See you in Torino🥪! @weidingerlaura @tvr2c @maurice_jks @GoogleDeepMind @Bauhaus_Uni
In the last year our PhD students @iLab_hwu achieved: finalists in Amazon Alexa Simbot challenge; Best Demo paper at EACL; papers at NAACL, HRI, NeurIPS, EMNLP, SIGDIAL, Interspeech … full list at sites.google.com/site/hwinterac…
Looking to do PhD in AI ? Our new CDT program in Dependable and Deployable AI for Robotics (CDT-D2AIR) offers funded PhDs & a supportive research environment @iLab_hwu @EDINrobotics @NRobotarium #GenAI #NLProc #HRI #Conversational_AI #PhD edinburgh-robotics.org/apply
At this week’s reading group, Eimear facilitated a discussion on the anthropomorphism of dialogue systems and other NLP tools, based on the paper "Mirages. On Anthropomorphism in Dialogue Systems." by @gavin_does_nlp @curriedamanda @t_dinkar @verena_rieser @ZeerakTalat
I have also been fortunate to count many women among my collaborators after my PhD: @albacurry @florplaza22 @t_dinkar @debora_nozza @donyarooein and my dear friend and co-host @orianajanem
Many cultures don't follow the Anglo-Saxon standard of single last names (e.g. John Smith) and at the moment the @aclanthology has issues assigning papers to people with two last names, leading some of us to have several profiles with our work scattered throughout them. 1/5
The CFP for the Third Workshop on Safety for Conversational AI @LRECCOLING is out. Please consider submitting to this workshop and join us in lovely Torino! @CurriedAmanda @peppeatta @sinantie @verena_rieser @dirk_hovy Full details here: sites.google.com/view/safety-co…
Also at @emnlpmeeting "Mirages. On Anthropomorphism in Dialogue Systems" by @gavin_does_nlp , @CurriedAmanda , @t_dinkar , @verena_rieser & @ZeerakTalat Anthropomorphism in dialogue systems arxiv.org/pdf/2305.09800…
Semantics at @semdialmeeting "Modelling Disagreement or Modelling Perspectives?" by @NikVits, @amitkparekh_, @t_dinkar, @gavin_does_nlp, @sinantie & @verena_rieser We predict disagreement on subjective data while preserving individual perspectives! aclanthology.org/2023.semeval-1…
An LBR at @ieeeroman "FurNav: Development and Preliminary Study of a Robot Direction Giver" by @bww_robotics, Yann Schlosser, Rayane Tarkany, Meriam Moujahid, @BirtheNesset, @t_dinkar & @verena_rieser Exploring direction giving styles @furhatrobotics arxiv.org/abs/2309.14499
This paper was a blast to write and a blast to present! @ZeerakTalat @verena_rieser @gavin_does_nlp @t_dinkar #EMNLP2023
Come check out our poster at #EMNLP2023! @t_dinkar @gavin_does_nlp
Thanks @CurriedAmanda for inviting me and for the candid picture!
For our weekly @MilaNLProc lab seminar, it was a pleasure to have @t_dinkar for a talk about Safety in Conversational AI. #NLProc
Yesterday, we presented our demo paper (virtually) at @sigdial.🥳 Thanks to: @talkwithfinny @ShabanaSheena, Nivan Nelson, Abhiram Karukayil, @addictedkingAK, Mohammed Afil Farhan, @BirtheNesset, Meriam Moujahid, @t_dinkar @verena_rieser @oliverlemon @NRobotarium