Jasmijn Bastings @ ACL 2025 πŸ‡¦πŸ‡Ή's Avatar

Jasmijn Bastings @ ACL 2025 πŸ‡¦πŸ‡Ή

@jasmijn.bastings.me

Senior Research Scientist at Google DeepMind. Interested in equitable language technology, gender, bias, interpretability, feminism, NLP. Views my own. She/her. 🌐 jasmijn.bastings.me

6,507 Followers  |  1,924 Following  |  7 Posts  |  Joined: 02.02.2024  |  1.6675

Latest posts by jasmijn.bastings.me on Bluesky

Amazing work by @mxeddie.bsky.social

04.08.2025 09:44 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Preview
Amplifying Trans and Nonbinary Voices: A Community-Centred Harm Taxonomy for LLMs Eddie L. Ungless, Sunipa Dev, Cynthia L. Bennett, Rebecca Gulotta, Jasmijn Bastings, Remi Denton. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Lon...

"Amplifying Trans and Nonbinary Voices: A Community-Centred Harm Taxonomy for LLMs" is online: aclanthology.org/2025.acl-lon... We explore LLM responses that may negatively impact the transgender and nonbinary (TGNB) community and introduce a proof-of-concept toolkit to help identify them #NLProc

04.08.2025 09:44 β€” πŸ‘ 13    πŸ” 2    πŸ’¬ 1    πŸ“Œ 0

@bsavoldi.bsky.social taking us back in time at #GITT2025 ⌚⏳ focusing on the first discussions of gender bias in language technology as a socio-technical issue. No, the problem hasn't been fixed yet. But what has happened?

23.06.2025 07:22 β€” πŸ‘ 7    πŸ” 3    πŸ’¬ 6    πŸ“Œ 0
Qualtrics Survey | Qualtrics Experience Management Survey of researchers in academia, industry and third sector who deploy LLMs as part of their work (e.g. because they evaluate LLMs, finetune LLMs, use LLMs to summarise data etc).

I've created a survey on ethics resources. Anyone in UK who does research with LLMs, in academia or industry, building models or using them as tools, is eligible. Takes 10 mins and there's the chance to win Β£100 voucher.

Survey: lnkd.in/eFRsqpT2

(V1 of the guide is here arxiv.org/abs/2410.19812)

23.05.2025 15:25 β€” πŸ‘ 4    πŸ” 5    πŸ’¬ 0    πŸ“Œ 0

My latest paper "A decade of gender bias in machine translation" with @bsavoldi.bsky.social @luisabentivogli.bsky.social and Eva Vanmassenhove is out. πŸ˜€β†˜οΈ #NLProc #NLP #MT

02.05.2025 18:33 β€” πŸ‘ 50    πŸ” 14    πŸ’¬ 1    πŸ“Œ 1

@evavnmssnhv.bsky.social

02.05.2025 18:34 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

My latest paper "A decade of gender bias in machine translation" with @bsavoldi.bsky.social @luisabentivogli.bsky.social and Eva Vanmassenhove is out. πŸ˜€β†˜οΈ #NLProc #NLP #MT

02.05.2025 18:33 β€” πŸ‘ 50    πŸ” 14    πŸ’¬ 1    πŸ“Œ 1
Preview
When ChatGPT Broke an Entire Field: An Oral History | Quanta Magazine Researchers in β€œnatural language processing” tried to tame human language. Then came the transformer.

This is a fantastic oral history of the last 10 years of NLP and AI. www.quantamagazine.org/when-chatgpt...

01.05.2025 11:55 β€” πŸ‘ 96    πŸ” 31    πŸ’¬ 2    πŸ“Œ 4
Preview
Not In Our Name: Feminist Academics and Educators Speak Out Against Transphobia We call on all trans-inclusive feminist academics and educators to sign this statement: β€˜Not In Our Name’ We are non-trans feminist academics and educators. We write in support of trans rights, tran...

If you are a feminist academic, you should sign this letter. forms.gle/oDYgnobrMiSc...

30.04.2025 07:12 β€” πŸ‘ 351    πŸ” 257    πŸ’¬ 13    πŸ“Œ 23
Barchart showing meta-llama/Meta-Llama-3-8B, meta-llama/Llama-3.1-70B, meta-llama/Llama-3.1-70B-Instruct. Each model has two bars, a blue one saying PRO-LEFT and a red one saying PRO-RIGHT. The PRO-RIGHT bars on higher than the PRO-LEFT in all of them.

Barchart showing meta-llama/Meta-Llama-3-8B, meta-llama/Llama-3.1-70B, meta-llama/Llama-3.1-70B-Instruct. Each model has two bars, a blue one saying PRO-LEFT and a red one saying PRO-RIGHT. The PRO-RIGHT bars on higher than the PRO-LEFT in all of them.

πŸ‘‰ πŸ‘ˆ Meta announced that they're changing their models to reduce "left-leaning [political] bias"--that means leaning them to the political "right". Lots to unpack about what that might mean. So I ran a quick "shot in the dark" study...and found a *political right* bias in Meta models. Some notes.🧡

22.04.2025 00:39 β€” πŸ‘ 711    πŸ” 276    πŸ’¬ 28    πŸ“Œ 37

This is a really good piece on why the UK supreme court ruling is so problematic.

22.04.2025 12:55 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Reserved topic scholarships | Doctoral Program - Information Engineering and Computer Science

πŸ“’ Come and join our group!
We offer a fully funded 3-year PhD position:

πŸ“” Automatic translation with large multimodal models: iecs.unitn.it/education/ad...

πŸ“Full details for application: iecs.unitn.it/education/ad...

πŸ“… Deadline May 12, 2025

#NLProc #FBK

22.04.2025 10:14 β€” πŸ‘ 10    πŸ” 10    πŸ’¬ 1    πŸ“Œ 0
Preview
The Supreme Court ignored trans voices. I’m ashamed of what our law has become | Good Law Project The Supreme Court’s decision to exclude trans voices has harmed trans people – and harmed the fairness of their ruling.

I've put on my hob nailed boots and given the Supreme Court decision the kicking it deserves. goodlaw.social/vd3a

16.04.2025 17:20 β€” πŸ‘ 955    πŸ” 361    πŸ’¬ 25    πŸ“Œ 32
Preview
On Fake Hannah Arendt Quotations This article discusses a mis-attributed quote to Hannah Arendt circulating on social media, exploring why such alterations occur and their potential consequences. Berkowitz argues that while simplifie...

I heard another fake Hannah Arendt quote on the radio. Since in the current times we might hear more of her to make sense of what is going on, here is a nice piece on what she actually said on lies & people not believing anything anymore: hac.bard.edu/amor-mundi/o...

12.01.2025 10:57 β€” πŸ‘ 8    πŸ” 0    πŸ’¬ 2    πŸ“Œ 0

My heartfelt condolences to everyone who knew and loved Felix. He was such a bright thinker. Such a loss.

03.01.2025 22:03 β€” πŸ‘ 3    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Post image

We scaled training data attribution (TDA) methods ~1000x to find influential pretraining examples for thousands of queries in an 8B-parameter LLM over the entire 160B-token C4 corpus!
medium.com/people-ai-re...

13.12.2024 18:57 β€” πŸ‘ 36    πŸ” 8    πŸ’¬ 2    πŸ“Œ 5
Screenshot of the paper.

Screenshot of the paper.

Even as an interpretable ML researcher, I wasn't sure what to make of Mechanistic Interpretability, which seemed to come out of nowhere not too long ago.

But then I found the paper "Mechanistic?" by
@nsaphra.bsky.social and @sarah-nlp.bsky.social, which clarified things.

20.11.2024 08:00 β€” πŸ‘ 232    πŸ” 28    πŸ’¬ 8    πŸ“Œ 2

How do LLMs learn to reason from data? Are they ~retrieving the answers from parametric knowledge🦜? In our new preprint, we look at the pretraining data and find evidence against this:

Procedural knowledge in pretraining drives LLM reasoning βš™οΈπŸ”’

πŸ§΅β¬‡οΈ

20.11.2024 16:31 β€” πŸ‘ 861    πŸ” 141    πŸ’¬ 38    πŸ“Œ 24

@jasmijn.bastings.me is following 20 prominent accounts