Timing Actions in Games Through Bio-Inspired Reinforcement Learning
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
A bio-inspired version of Reinforcement Learning (RL) can be used to learn to plan actions in a fully neuromorphic robot, allowing perception, processing, action planning, and control, maintaining an end-to-end spiking signal. Such an agent could fully take advantage of the sparse, low-power encoding and give insight into the secrets of biological intelligence. The current state-of-the-art in neuromorphic RL uses populations of neurons to implement traditional RL equations with novel spiking state-representation methods and achieves learning through weight updates of neural connections in an 8 × 8 grid world with discrete state definitions. We adapt and extend the algorithm towards a fully neuromorphic robot capable of playing highly dynamic games. In this paper we integrate the RL algorithm with a robot simulator for air hockey; doubling the dimensionality of the, now continuous, state. We demonstrate that we can adapt the method to learn precise ‘hit timing’, as the puck moves in front of the robot, the robot must choose the correct timing to intercept the puck, knocking it towards the opponent's goal. We also introduce a developmental approach to learning with Curriculum Learning (CL), allowing the robot to first learn a simple task, which can then be generalised and refined to more complex scenarios. The simplified air-hockey scenario demonstrates promising results for a fully neuromorphic pipeline in the future.
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.
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 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 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 it