Matches in SemOpenAlex for { <https://semopenalex.org/work/W4319453300> ?p ?o ?g. }
Showing items 1 to 67 of
67
with 100 items per page.
- W4319453300 abstract "Learning from human preferences is important for language models to match human needs and to align with human and social values. Prior works have achieved remarkable successes by learning from human feedback to understand and follow instructions. Nonetheless, these methods are either founded on hand-picked model generations that are favored by human annotators, rendering them inefficient in terms of data utilization and challenging to apply in general, or they depend on reinforcement learning, which often suffers from imperfect reward functions and relies on extremely challenging optimizations. In this work, we propose a novel technique, Chain of Hindsight, that is easy to optimize and can learn from any form of feedback, regardless of its polarity. Our idea is inspired by how humans learn from extensive feedback presented in the form of languages. We convert all types of feedback into sequences of sentences, which are then used to fine-tune the model, allowing us to take advantage of the language comprehension capabilities of language models. We condition the model on a sequence of model generations paired with feedback. By doing so, the model is trained to generate outputs based on feedback, while learning to identify and correct negative attributes or errors. Applying our method to large language models, we observed that Chain of Hindsight significantly surpasses previous methods in aligning language models with human preferences. We report significant improvements on summarization and dialogue benchmarks, with our approach markedly preferred in human evaluations." @default.
- W4319453300 created "2023-02-09" @default.
- W4319453300 creator A5041119232 @default.
- W4319453300 creator A5049349154 @default.
- W4319453300 creator A5088664989 @default.
- W4319453300 date "2023-02-06" @default.
- W4319453300 modified "2023-10-06" @default.
- W4319453300 title "Chain of Hindsight Aligns Language Models with Feedback" @default.
- W4319453300 doi "https://doi.org/10.48550/arxiv.2302.02676" @default.
- W4319453300 hasPublicationYear "2023" @default.
- W4319453300 type Work @default.
- W4319453300 citedByCount "0" @default.
- W4319453300 crossrefType "posted-content" @default.
- W4319453300 hasAuthorship W4319453300A5041119232 @default.
- W4319453300 hasAuthorship W4319453300A5049349154 @default.
- W4319453300 hasAuthorship W4319453300A5088664989 @default.
- W4319453300 hasBestOaLocation W43194533001 @default.
- W4319453300 hasConcept C10347200 @default.
- W4319453300 hasConcept C119857082 @default.
- W4319453300 hasConcept C137293760 @default.
- W4319453300 hasConcept C138885662 @default.
- W4319453300 hasConcept C154945302 @default.
- W4319453300 hasConcept C15744967 @default.
- W4319453300 hasConcept C170858558 @default.
- W4319453300 hasConcept C180747234 @default.
- W4319453300 hasConcept C204321447 @default.
- W4319453300 hasConcept C205711294 @default.
- W4319453300 hasConcept C2778112365 @default.
- W4319453300 hasConcept C2780310539 @default.
- W4319453300 hasConcept C41008148 @default.
- W4319453300 hasConcept C41895202 @default.
- W4319453300 hasConcept C54355233 @default.
- W4319453300 hasConcept C86803240 @default.
- W4319453300 hasConcept C97541855 @default.
- W4319453300 hasConceptScore W4319453300C10347200 @default.
- W4319453300 hasConceptScore W4319453300C119857082 @default.
- W4319453300 hasConceptScore W4319453300C137293760 @default.
- W4319453300 hasConceptScore W4319453300C138885662 @default.
- W4319453300 hasConceptScore W4319453300C154945302 @default.
- W4319453300 hasConceptScore W4319453300C15744967 @default.
- W4319453300 hasConceptScore W4319453300C170858558 @default.
- W4319453300 hasConceptScore W4319453300C180747234 @default.
- W4319453300 hasConceptScore W4319453300C204321447 @default.
- W4319453300 hasConceptScore W4319453300C205711294 @default.
- W4319453300 hasConceptScore W4319453300C2778112365 @default.
- W4319453300 hasConceptScore W4319453300C2780310539 @default.
- W4319453300 hasConceptScore W4319453300C41008148 @default.
- W4319453300 hasConceptScore W4319453300C41895202 @default.
- W4319453300 hasConceptScore W4319453300C54355233 @default.
- W4319453300 hasConceptScore W4319453300C86803240 @default.
- W4319453300 hasConceptScore W4319453300C97541855 @default.
- W4319453300 hasLocation W43194533001 @default.
- W4319453300 hasOpenAccess W4319453300 @default.
- W4319453300 hasPrimaryLocation W43194533001 @default.
- W4319453300 hasRelatedWork W1492315459 @default.
- W4319453300 hasRelatedWork W2139970489 @default.
- W4319453300 hasRelatedWork W2540910169 @default.
- W4319453300 hasRelatedWork W2993601805 @default.
- W4319453300 hasRelatedWork W3089780453 @default.
- W4319453300 hasRelatedWork W3140454661 @default.
- W4319453300 hasRelatedWork W3148904318 @default.
- W4319453300 hasRelatedWork W3197854638 @default.
- W4319453300 hasRelatedWork W4245029315 @default.
- W4319453300 hasRelatedWork W4319453300 @default.
- W4319453300 isParatext "false" @default.
- W4319453300 isRetracted "false" @default.
- W4319453300 workType "article" @default.