Better Time Constrained Search via Randomization and Postprocessing
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
Most of the satisficing planners which are based on heuristic search iteratively improve their solution quality through an anytime approach. Typically, the lowest-cost solution found so far is used to constrain the search. This avoids areas of the state space which cannot directly lead to lower cost solutions. However, in this paper we show that when used in conjunction with a post-processing plan improvement system such as ARAS, this bounding approach can harm a planner’s performance since the bound may prevent the search from ever finding additional plans for the post-processor to improve.The new anytime search framework of Diverse Any-Time Search addresses this issue through the use of restarts, randomization, and by not bounding as strictly as is done by previous approaches. Below, we will show that by using these techniques, the framework is able to generate a more diverse set of “raw" input plans for the post-processor to work on. We then show that when adding both Diverse Any-Time Search and the ARAS post-processor to LAMA-2011, the winner of the most recent IPC planning competition, the performance according to the IPC scoring metric improves from 511 points to over 570 points when tested on the 550 problems from IPC 2008 and IPC 2011. Performance gains are also seen when these techniques are added to Anytime Explicit Estimation Algorithm (AEES), as the performance improves from 440 points to over 513 points on the same problem set.
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 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 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