Evaluating real-time strategy game states using convolutional neural networks
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
Real-time strategy (RTS) games, such as Blizzard's StarCraft, are fast paced war simulation games in which players have to manage economies, control many dozens of units, and deal with uncertainty about opposing unit locations in real-time. Even in perfect information settings, constructing strong AI systems has been difficult due to enormous state and action spaces and the lack of good state evaluation functions and high-level action abstractions. To this day, good human players are still handily defeating the best RTS game AI systems, but this may change in the near future given the recent success of deep convolutional neural networks (CNNs) in computer Go, which demonstrated how networks can be used for evaluating complex game states accurately and to focus look-ahead search. In this paper we present a CNN for RTS game state evaluation that goes beyond commonly used material based evaluations by also taking spatial relations between units into account. We evaluate the CNN's performance by comparing it with various other evaluation functions by means of tournaments played by several state-of-the-art search algorithms. We find that, despite its much slower evaluation speed, on average the CNN based search performs significantly better compared to simpler but faster evaluations. These promising initial results together with recent advances in hierarchical search suggest that dominating human players in RTS games may not be far off.
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 | 0.001 | 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 | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 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