Eoin Kenny (@eoinknny) 's Twitter Profile
Eoin Kenny

@eoinknny

Senior Associate AI Researcher @ JPMC
🇮🇪 I am interested in how interpretable ML and humans can work together.

ID: 426256713

linkhttps://eoinkenny.github.io/ calendar_today02-12-2011 01:12:30

179 Tweet

141 Followers

85 Following

Eoin Kenny (@eoinknny) 's Twitter Profile Photo

Our NeurIPS paper is now available in preprint here arxiv.org/abs/2310.18937 We're interested in exploring different types of contrastive explanation, and show counterfactuals are not always the more useful option. #NeurIPS2023 #XAI #MachineLearning

Our NeurIPS paper is now available in preprint here arxiv.org/abs/2310.18937

We're interested in exploring different types of contrastive explanation, and show counterfactuals are not always the more useful option.

#NeurIPS2023 #XAI #MachineLearning
Eoin Kenny (@eoinknny) 's Twitter Profile Photo

'The utility of "even if..." semifactual explanation to optimize positive outcomes' Happy to post the presentation for our NeurIPS paper :recorder-v3.slideslive.com/?share=86709&s… Please share to those interested in #causal #explainability and #algorithmic #recourse in #ethicalML and #XAI. 😀🎉

Andi Peng (@theandipenguin) 's Twitter Profile Photo

Humans use abstractions for data-efficient learning. We wish for neural networks to do the same. In our proposed human-in-the-loop framework, we automatically generate a spectrum of abstractions and allow users to deploy task-appropriate ones. To appear at #NeurIPS2023! [1/n]

Humans use abstractions for data-efficient learning. We wish for neural networks to do the same. 

In our proposed human-in-the-loop framework, we automatically generate a spectrum of abstractions and allow users to deploy task-appropriate ones. 

To appear at #NeurIPS2023! [1/n]
Giang Nguyen (@giangnguyen2412) 's Twitter Profile Photo

We had fun and passionate discussion today at the XAI meetup @NeurIPS2023. A lot of new/old faces in the field. Don't forget to follow each other for future meetups/possible collaboration 🙌🙌🙌

We had fun and passionate discussion today at the XAI meetup @NeurIPS2023. A lot of new/old faces in the field. Don't forget to follow each other for future meetups/possible collaboration 🙌🙌🙌
Jinay (@jinaycodes) 's Twitter Profile Photo

turns out the project I built to learn proximal policy optimization was used in a notable ICLR 2023 paper on RL interpretability thank you for the honor Eoin Kenny, Mycal Tucker, and Julie Shah!

turns out the project I built to learn proximal policy optimization was used in a notable ICLR 2023 paper on RL interpretability

thank you for the honor <a href="/EoinKNNy/">Eoin Kenny</a>, <a href="/MycalTucker/">Mycal Tucker</a>, and <a href="/julie_a_shah/">Julie Shah</a>!
Explainable AI (@xai_research) 's Twitter Profile Photo

For those that weren’t able to make it to Eoin’s talk, please check it out on our XAI Youtube page! youtube.com/watch?v=8oPI7W…

Eoin Kenny (@eoinknny) 's Twitter Profile Photo

This new DeepSeek LLM has totally changed everything, why pay OpenAI massive prices for a slow GPT-4o when a comparably free version exists that is even better and open source? Perhaps academia can now finally enter the AI arms race.

Eoin Kenny (@eoinknny) 's Twitter Profile Photo

I have 8 papers to review for ICML, I'm sincerely sorry to those authors I might review, I will try to do my best but 2 papers a week along with everything in my personal and professional life will be hard to do well.

Eoin Kenny (@eoinknny) 's Twitter Profile Photo

Responding to reviewers at ICML has become an exercise in how to do a rebuttal against an LLM looking for nonsense critiques. "Oh the paper only evaluated on 10 datasets, I reject everything with < 100", good bye and good luck human!