Uncertainty aware behavioral cloning using Bayesian Neural Networks
Bibliographic record
Abstract
Leveraging demonstrations to learn complex maneuvers is a handy technique to teach robots when other techniques like reinforcement learning are difficult to implement.However, the world is diverse and partially observable, and generating demonstrations under all possible situations in hard.As a result, a lot of times the robot's decision on the situation it has not seen earlier is wrong.One line of potential solutions to handle this problem of incorrect decision-making at previously unseen situations begins with equipping the robots with a sense of confidence in their decisions.In this work, we develop one such technique using a data-driven probabilistic method.Specifically, we use a Bayesian Neural Network to generate a cheap to obtain quantity for the robot to gauge its confidence of doing well in a given situation.These different situations are more relatable to the real world in the sense that the differences between them are unobservable.We show the scalability and consistency of our approach to such situations in high dimensional simulated and real robotic domains.Also, we show that such a confidence based solution allows making an informed decision about when to invoke a fallback strategy.One fallback strategy is to request more data.We empirically show that providing data only when requested results in increased data-efficiency.This is crucial in the real-world as data is expensive and painstaking to obtain.I would first like to thank my first co-supervisor Dr. Doina Precup for allowing me to be part of the RLLab and for her constant support throughout during the duration of my thesis.I would deeply thank my second co-supervisor Dr. David Meger for allowing me the opportunity to work in the Mobile Robotics Lab and for his constant guidance, and feedback on my work.I am also grateful to him for the amount of time he had taken out of his schedule weekly and ease me through the challenging process of doing a fruitful and constructive work.Dr. David Meger
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 0.000 |
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.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".