Razvan Pascanu, Maciej Wołczyk, Łukasz Kuci, Piotr Miłoś, Michał Zaj Ąc, Mateusz Ostaszewski, Bartłomiej Cupiał, Michał Bortkiewicz
We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Fine-tuning is a widespread technique that allows practitioners to transfer pre-trained capabilities, as recently showcased by the successful applications of foundation models. However, fine-tuning reinforcement learning (RL) models remains a challenge. This work conceptualizes one specific cause of poor transfer, accentuated in the RL setting by the interplay between actions and observations: forgetting of pre-trained capabilities. Namely, a model deteriorates on the state subspace of the downstream task not visited in the initial phase of fine-tuning, on which the model behaved well due to pre-training. This way, we lose the anticipated transfer benefits. We identify conditions when this problem occurs, showing that it is common and, in many cases, catastrophic. Through a detailed empirical analysis of the challenging NetHack and Montezuma's Revenge environments, we show that standard knowledge retention techniques mitigate the problem and thus allow us to take full advantage of the pre-trained capabilities. In particular, in NetHack, we achieve a new state-of-the-art for neural models, improving the previous best score from 5K to over 10K points in the Human Monk scenario.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2402.02868")
get_code_for_paper("2402.02868")
have("2402.02868")
The run record, dated, one paper per request, free:
curl https://syntology.ai/api/ran/2402.02868.json
A badge for a README (the split and the date, never a ratio):
[](https://syntology.ai/paper/2402.02868)
Connect an agent — have() is free.