Dang Nguyen
@divingwithorcas.bsky.social
Computer Science PhD student at UChicago | Member of the Chicago Human+AI lab @chicagohai.bsky.social
Paper: arxiv.org/abs/2504.06303
On the Effectiveness and Generalization of Race Representations for Debiasing High-Stakes Decisions
Understanding and mitigating biases is critical for the adoption of large language models (LLMs) in high-stakes decision-making. We introduce Admissions and Hiring, decision tasks with hypothetical ap...
arxiv.org
October 8, 2025 at 3:14 PM
Paper: arxiv.org/abs/2504.06303
and yes, you can play on your mobile browser: hrsimulator.communicationgames.ai
HR Simulator™: Be the Person You Hate
A game that will change how you write emails.
hrsimulator.communicationgames.ai
September 29, 2025 at 2:50 AM
and yes, you can play on your mobile browser: hrsimulator.communicationgames.ai
Please use a VPN. We're sorry for any inconvenience!
September 27, 2025 at 2:26 PM
Please use a VPN. We're sorry for any inconvenience!
Stay tuned for more on communication games! Big thanks to @ari-holtzman.bsky.social @Harvey Fu @chenhaotan.bsky.social @Peter West for making this project happen!
September 26, 2025 at 6:48 PM
Stay tuned for more on communication games! Big thanks to @ari-holtzman.bsky.social @Harvey Fu @chenhaotan.bsky.social @Peter West for making this project happen!
hrsimulator.communicationgames.ai
We’re serious! Economic coordination happens via emails. How do humans fare against AIs in getting things done with words?
We see a genre co-emerging with LLMs: communication games, where communication is crucial and not just “cheap talk” like Mafia or Diplomacy.
We’re serious! Economic coordination happens via emails. How do humans fare against AIs in getting things done with words?
We see a genre co-emerging with LLMs: communication games, where communication is crucial and not just “cheap talk” like Mafia or Diplomacy.
HR Simulator™: Be the Person You Hate
A game that will change how you write emails.
hrsimulator.communicationgames.ai
September 26, 2025 at 6:45 PM
hrsimulator.communicationgames.ai
We’re serious! Economic coordination happens via emails. How do humans fare against AIs in getting things done with words?
We see a genre co-emerging with LLMs: communication games, where communication is crucial and not just “cheap talk” like Mafia or Diplomacy.
We’re serious! Economic coordination happens via emails. How do humans fare against AIs in getting things done with words?
We see a genre co-emerging with LLMs: communication games, where communication is crucial and not just “cheap talk” like Mafia or Diplomacy.
12/n
Big thanks to @chenhaotan.bsky.social for advice on the project, as well as helpful feedback from the wonderful members of the @chicagohai.bsky.social lab! Check out our code at github.com/ChicagoHAI/l....
DM me for any questions!
Big thanks to @chenhaotan.bsky.social for advice on the project, as well as helpful feedback from the wonderful members of the @chicagohai.bsky.social lab! Check out our code at github.com/ChicagoHAI/l....
DM me for any questions!
GitHub - ChicagoHAI/llm-prediction-bias
Contribute to ChicagoHAI/llm-prediction-bias development by creating an account on GitHub.
github.com
April 14, 2025 at 8:13 PM
12/n
Big thanks to @chenhaotan.bsky.social for advice on the project, as well as helpful feedback from the wonderful members of the @chicagohai.bsky.social lab! Check out our code at github.com/ChicagoHAI/l....
DM me for any questions!
Big thanks to @chenhaotan.bsky.social for advice on the project, as well as helpful feedback from the wonderful members of the @chicagohai.bsky.social lab! Check out our code at github.com/ChicagoHAI/l....
DM me for any questions!
11/n
So strangely, changing the prompt can change how a model represents race. Thus, in some cases, the model’s representation may be sensitive to spurious prompt features, which poses a challenge to the generalizability of debiasing methods. Future work on debiasing should take this into account.
So strangely, changing the prompt can change how a model represents race. Thus, in some cases, the model’s representation may be sensitive to spurious prompt features, which poses a challenge to the generalizability of debiasing methods. Future work on debiasing should take this into account.
April 14, 2025 at 8:12 PM
11/n
So strangely, changing the prompt can change how a model represents race. Thus, in some cases, the model’s representation may be sensitive to spurious prompt features, which poses a challenge to the generalizability of debiasing methods. Future work on debiasing should take this into account.
So strangely, changing the prompt can change how a model represents race. Thus, in some cases, the model’s representation may be sensitive to spurious prompt features, which poses a challenge to the generalizability of debiasing methods. Future work on debiasing should take this into account.
10/n
We found the race subspace generalizes cross-family (from admissions to hiring) and, to a lesser extent, cross-explicitness (from implicit race via name to explicit race), but it fails to generalize cross-prompt (from one prompt template to another).
We found the race subspace generalizes cross-family (from admissions to hiring) and, to a lesser extent, cross-explicitness (from implicit race via name to explicit race), but it fails to generalize cross-prompt (from one prompt template to another).
April 14, 2025 at 8:11 PM
10/n
We found the race subspace generalizes cross-family (from admissions to hiring) and, to a lesser extent, cross-explicitness (from implicit race via name to explicit race), but it fails to generalize cross-prompt (from one prompt template to another).
We found the race subspace generalizes cross-family (from admissions to hiring) and, to a lesser extent, cross-explicitness (from implicit race via name to explicit race), but it fails to generalize cross-prompt (from one prompt template to another).