Andrea Pugnana
@andrepugni.bsky.social
Research Fellow @ University of Trento. Studying ML models that know what they do not know. He/Him.
For other believers, link to the old arxiv, arxiv.org/abs/2503.161.... updated version to be out soon.
Deferring Concept Bottleneck Models: Learning to Defer Interventions to Inaccurate Experts
Concept Bottleneck Models (CBMs) are machine learning models that improve interpretability by grounding their predictions on human-understandable concepts, allowing for targeted interventions in their...
arxiv.org
September 18, 2025 at 9:39 PM
For other believers, link to the old arxiv, arxiv.org/abs/2503.161.... updated version to be out soon.
👀 Want to dive deeper? Curious how ML + humans really perform together?
📍Come find us at our poster: Hall A-E 160, May 3rd.
5/5
📍Come find us at our poster: Hall A-E 160, May 3rd.
5/5
April 29, 2025 at 9:42 PM
👀 Want to dive deeper? Curious how ML + humans really perform together?
📍Come find us at our poster: Hall A-E 160, May 3rd.
5/5
📍Come find us at our poster: Hall A-E 160, May 3rd.
5/5
🔍 We define wo scenarios:
1️⃣ Both human and ML predictions for deferred cases? We can estimate individual & aggregate causal effects.
2️⃣ Only human predictions for deferred cases? We can use Regression Discontinuity Design for local causal effects.
4/5
1️⃣ Both human and ML predictions for deferred cases? We can estimate individual & aggregate causal effects.
2️⃣ Only human predictions for deferred cases? We can use Regression Discontinuity Design for local causal effects.
4/5
April 29, 2025 at 9:42 PM
🔍 We define wo scenarios:
1️⃣ Both human and ML predictions for deferred cases? We can estimate individual & aggregate causal effects.
2️⃣ Only human predictions for deferred cases? We can use Regression Discontinuity Design for local causal effects.
4/5
1️⃣ Both human and ML predictions for deferred cases? We can estimate individual & aggregate causal effects.
2️⃣ Only human predictions for deferred cases? We can use Regression Discontinuity Design for local causal effects.
4/5
In our work, we bring causal inference to the arena. Using the potential outcomes framework, we ask: What’s the actual impact of deferring to humans on system performance? 🎯
3/5
3/5
April 29, 2025 at 9:42 PM
In our work, we bring causal inference to the arena. Using the potential outcomes framework, we ask: What’s the actual impact of deferring to humans on system performance? 🎯
3/5
3/5
🧠🤖 Deferring systems let ML models pass the mic to humans when they are better at predicting. But what is the actual impact on predictive performance of such systems? Surprisingly, such a question is still overlooked.
2/5
2/5
April 29, 2025 at 9:42 PM
🧠🤖 Deferring systems let ML models pass the mic to humans when they are better at predicting. But what is the actual impact on predictive performance of such systems? Surprisingly, such a question is still overlooked.
2/5
2/5
👀 Want to dive deeper? Curious how ML + humans really perform together?
📍Come find us at our poster: Hall A-E 160, May 3rd.
5/5
📍Come find us at our poster: Hall A-E 160, May 3rd.
5/5
April 29, 2025 at 9:37 PM
👀 Want to dive deeper? Curious how ML + humans really perform together?
📍Come find us at our poster: Hall A-E 160, May 3rd.
5/5
📍Come find us at our poster: Hall A-E 160, May 3rd.
5/5
🔍 We define wo scenarios:
1️⃣ Both human and ML predictions for deferred cases? We can estimate individual & aggregate causal effects.
2️⃣ Only human predictions for deferred cases? We can use Regression Discontinuity Design for local causal effects.
4/5
1️⃣ Both human and ML predictions for deferred cases? We can estimate individual & aggregate causal effects.
2️⃣ Only human predictions for deferred cases? We can use Regression Discontinuity Design for local causal effects.
4/5
April 29, 2025 at 9:37 PM
🔍 We define wo scenarios:
1️⃣ Both human and ML predictions for deferred cases? We can estimate individual & aggregate causal effects.
2️⃣ Only human predictions for deferred cases? We can use Regression Discontinuity Design for local causal effects.
4/5
1️⃣ Both human and ML predictions for deferred cases? We can estimate individual & aggregate causal effects.
2️⃣ Only human predictions for deferred cases? We can use Regression Discontinuity Design for local causal effects.
4/5
In our work, we bring causal inference to the arena. Using the potential outcomes framework, we ask: What’s the actual impact of deferring to humans on system performance? 🎯
3/5
3/5
April 29, 2025 at 9:37 PM
In our work, we bring causal inference to the arena. Using the potential outcomes framework, we ask: What’s the actual impact of deferring to humans on system performance? 🎯
3/5
3/5
🧠🤖 Deferring systems let ML models pass the mic to humans when they are better at predicting. But what is the actual impact on predictive performance of such systems? Surprisingly, such a question is still overlooked.
2/5
2/5
April 29, 2025 at 9:37 PM
🧠🤖 Deferring systems let ML models pass the mic to humans when they are better at predicting. But what is the actual impact on predictive performance of such systems? Surprisingly, such a question is still overlooked.
2/5
2/5
Also quite curious to see what new technical issue will appear on Friday.
March 12, 2025 at 5:23 PM
Also quite curious to see what new technical issue will appear on Friday.
Finally, we thank the reviewers for their valuable suggestions and will add the link to the complete version as soon as possible. See you in Mai Khao! 4/4
January 22, 2025 at 11:29 AM
Finally, we thank the reviewers for their valuable suggestions and will add the link to the complete version as soon as possible. See you in Mai Khao! 4/4
Moreover, our experiments showcase how one can reason about these causal effects.
We believe that shifting the paradigm from pure performance to the causal effect of introducing a human-AI team is a necessary building block to make AI truly human-centred and aligned. 3/4
We believe that shifting the paradigm from pure performance to the causal effect of introducing a human-AI team is a necessary building block to make AI truly human-centred and aligned. 3/4
January 22, 2025 at 11:29 AM
Moreover, our experiments showcase how one can reason about these causal effects.
We believe that shifting the paradigm from pure performance to the causal effect of introducing a human-AI team is a necessary building block to make AI truly human-centred and aligned. 3/4
We believe that shifting the paradigm from pure performance to the causal effect of introducing a human-AI team is a necessary building block to make AI truly human-centred and aligned. 3/4