ConGolog, Sin Trans: compiling ConGolog into basic action theories for planning and beyond
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
ConGolog is a logical programming language for agents that is de-fined in the situation calculus. ConGolog agent control programs were originally proposed as an alternative to planning, but have also more re-cently been proposed as a means of providing domain control knowledge for planning. In this paper, we present a compiler that takes a ConGolog program and produces a new basic action theory of the situation calculus whose executable situations are all and only those that are permitted by the program. The size of the resulting theory is quadratic in the size of the original program – even in the face of unbounded loops, recursion, and concurrency. The compilation is of both theoretical and practical interest. From a theoretical perspective, proving properties of ConGolog programs is simplified because reification of programs is no longer required, and the compiled theory contains fewer second-order axioms. Further, in some cases, properties can be proven by regressing the program to the initial situation, eliminating the need for a higher order theorem prover. From a practical perspective, the compilation provides the mathematical founda-tion for compiling ConGolog programs into classical planning problems, including, with minor restrictions, into the Plan Domain Definition Lan-guage (PDDL), which is used as the input language for most state-of-the-art planners. Moreover, Hierarchical Task Networks (HTNs), a popular planning paradigm for industrial applications can be represented as Con-Golog programs and can thus now also be compiled to a classical planning problem. Such compilations are significant because they allow the best state-of-the-art planners to exploit ConGolog and HTN search control, without the need for special-purpose machinery.
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 | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| 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