Bibliographic record
Abstract
A tabular Reinforcement Learning algorithm, such as Q-learning, is incapable of solving problems with large state spaces due to explosive growth of the table required to be kept in fast memory (e.g. RAM). Neural Networks offer a solution at the expense of representation, among other things. Yet if we analyze the learning task in terms of the goal arity, we can approach problems of larger state spaces with Discretized Q-learning algorithm, by dramatically reducing memory requirements for the Q table.This paper demonstrates the methodology using the well-known LightsOut puzzle. While analytical methods of solving some variants of this puzzle are known, we proceed under a different premise of putting the learning agent in a much more difficult position of having virtually no initial information about the rules of the puzzle and often not being able to perceive the symmetries inherent in the two-dimensional puzzle grid.Without the described savings a standard and already thrifty Q-learning solution to a 5×5 variant of the puzzle would require over 3 gigabytes of RAM with 4 bytes per table entry, while our Discretized Q-learning solution required only 5 bits per table entry, allowing us to solve the puzzle on a 1 gigabyte machine. Finally, we also consider a variant of this puzzle for which no analytical solutions are known. The Discretized Q-learning algorithm is equally applicable to the new puzzle and can solve it with exactly the same effort.
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 imitationNot 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.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".