Philip Osborne
@philiposborne.bsky.social
Improving Reinforcement Learning with language interaction and humans in the loop.
PhD in Artificial Intelligence, University of Manchester (UK)
Founder of elsci.org
PhD in Artificial Intelligence, University of Manchester (UK)
Founder of elsci.org
I come here to escape 'this guy' 😂 I wouldn't be surprised if he set this up like the 'guardrails' you see on other llms just to generate the response he wanted
February 16, 2025 at 9:19 PM
I come here to escape 'this guy' 😂 I wouldn't be surprised if he set this up like the 'guardrails' you see on other llms just to generate the response he wanted
Really interesting work! In the conclusion for the ML community, is the idea that this could be used as a benchmark because you could introduce an agent (different from the investor/company agents) that optimises the esg policy rather than using a fixed comparison shown in the paper?
February 16, 2025 at 3:23 PM
Really interesting work! In the conclusion for the ML community, is the idea that this could be used as a benchmark because you could introduce an agent (different from the investor/company agents) that optimises the esg policy rather than using a fixed comparison shown in the paper?