MétaCan
Menu
Back to cohort
Record W4399552717 · doi:10.1613/jair.1.15155

Mitigating Value Hallucination in Dyna-Style Planning via Multistep Predecessor Models

2024· article· en· W4399552717 on OpenAlexaff
Farzane Aminmansour, Taher Jafferjee, Ehsan Imani, Erin J. Talvitie, Michael Bowling, Martha White

Bibliographic record

VenueJournal of Artificial Intelligence Research · 2024
Typearticle
Languageen
FieldComputer Science
TopicExplainable Artificial Intelligence (XAI)
Canadian institutionsUniversity of Alberta
Fundersnot available
KeywordsHallucinatingComputer scienceReinforcement learningBootstrapping (finance)Artificial intelligenceValue (mathematics)Function (biology)Successor cardinalSample (material)Bellman equationMachine learningAlgorithmMathematical optimizationEconometricsMathematics

Abstract

fetched live from OpenAlex

Dyna-style reinforcement learning (RL) agents improve sample efficiency over model-free RL agents by updating the value function with simulated experience generated by an environment model. However, it is often difficult to learn accurate models of environment dynamics, and even small errors may result in failure of Dyna agents. In this paper, we highlight that one potential cause of that failure is bootstrapping off of the values of simulated states, and introduce a new Dyna algorithm to avoid this failure. We discuss a design space of Dyna algorithms, based on using successor or predecessor models---simulating forwards or backwards---and using one-step or multi-step updates. Three of the variants have been explored, but surprisingly the fourth variant has not: using predecessor models with multi-step updates. We present the \emph{Hallucinated Value Hypothesis} (HVH): updating the values of real states towards values of simulated states can result in misleading action values which adversely affect the control policy. We discuss and evaluate all four variants of Dyna amongst which three update real states toward simulated states --- so potentially toward hallucinated values --- and our proposed approach, which does not. The experimental results provide evidence for the HVH, and suggest that using predecessor models with multi-step updates is a fruitful direction toward developing Dyna algorithms that are more robust to model error.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.010
metaresearch head score (Gemma)0.001
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesScholarly communication
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.906
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0100.001
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0020.003
Science and technology studies0.0000.000
Scholarly communication0.0010.004
Open science0.0020.000
Research integrity0.0000.002
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.199
GPT teacher head0.445
Teacher spread0.247 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one teacher head, not a consensus.

Study designSimulation or modeling
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2024
Admission routes1
Has abstractyes

Explore more

Same venueJournal of Artificial Intelligence ResearchSame topicExplainable Artificial Intelligence (XAI)French-language works237,207