Zobrazeno 1 - 5
of 5
pro vyhledávání: '"Balazadeh, Vahid"'
Autor:
Lau, Allison, Choi, Younwoo, Balazadeh, Vahid, Chidambaram, Keertana, Syrgkanis, Vasilis, Krishnan, Rahul G.
Reinforcement Learning from Human Feedback (RLHF) is widely used to align Language Models (LMs) with human preferences. However, existing approaches often neglect individual user preferences, leading to suboptimal personalization. We present the Pref
Externí odkaz:
http://arxiv.org/abs/2410.14001
Autor:
Balazadeh, Vahid, Chidambaram, Keertana, Nguyen, Viet, Krishnan, Rahul G., Syrgkanis, Vasilis
We study the problem of online sequential decision-making given auxiliary demonstrations from experts who made their decisions based on unobserved contextual information. These demonstrations can be viewed as solving related but slightly different ta
Externí odkaz:
http://arxiv.org/abs/2404.07266
Estimating the causal structure of observational data is a challenging combinatorial search problem that scales super-exponentially with graph size. Existing methods use continuous relaxations to make this problem computationally tractable but often
Externí odkaz:
http://arxiv.org/abs/2308.07480
We consider the problem of partial identification, the estimation of bounds on the treatment effects from observational data. Although studied using discrete treatment variables or in specific causal graphs (e.g., instrumental variables), partial ide
Externí odkaz:
http://arxiv.org/abs/2210.08139
Reinforcement learning agents have been mostly developed and evaluated under the assumption that they will operate in a fully autonomous manner -- they will take all actions. In this work, our goal is to develop algorithms that, by learning to switch
Externí odkaz:
http://arxiv.org/abs/2002.04258