Zayden-Jung/lean-bench
2
1
Fork 0
Code Issues 2 Pull Requests Actions Packages Projects Releases Wiki Activity

feat(schema): persist and instantiate scoped multi-role relation schemas
ci / python (pull_request) Successful in 17s
Details
ci / lean (pull_request) Successful in 1m6s
Details
ci / image (pull_request) Has been skipped
Details

Browse Source
This commit is contained in:
Codex Bot
2026-10-28 21:26:55 +08:00
parent 7d2c6f470b
commit 29fe3e1aa5
10 changed files with 2068 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/research-plan.md
+9 -3
View File
@@ -8,9 +8,15 @@ Evidence base: repo README and code (`leanbench/`, 4,345 lines incl. tests), pro
## 0. The idea in one page
**Terminology clarification (2026-10-28).** The owner uses "simplex" for a semantic,
multi-role relation schema: one statement joins objects, assumptions and a conclusion.
It does not mean a geometric simplex or assert a mathematical simplicial complex.
See [the implemented schema design](simplex-schema.md) for scoped slots, storage and
candidate retrieval. Existing signature projections and AI source annotations remain partial.
**Simplex, as it exists in the archived code.**
- `Role(name, context, choices)`: a named semantic slot with a description and an optional closed set of allowed values (`is_constrained`, `validate_value`).
- `Simplex(name, roles: {alias: Role})`: a reusable unit of **at least two** roles. A Simplex is itself a `Role`, so simplices nest; nesting gives a *simplicial complex*. `Equivalence` is the two-role special case.
- `Simplex(name, roles: {alias: Role})`: a reusable unit of **at least two** roles. A Simplex is itself a `Role`, so explicitly supplied schemas can be nested as semantic composition. `Equivalence` is the two-role special case.
- `State` / `Fact` / `Goal`: a role plus a value. A `Fact` has a value, a `Goal` does not; `SolvingTask.is_completed()` is "every goal resolved".
- `SchemaRegistry` (SQLite): `set_role`, `set_simplex`, `search_roles`, `find_by_roles(known_roles)`, `find_solvable(known_role_names)`.
- `SchemaParserAgent`: turns a query into a Simplex through a forced tool call whose JSON schema says `minItems: 2`, with a **search-first, reuse-before-create** rule for roles. `AnalogySchemaMiner`: proposes new simplices by the ratio *Simplex B : Role B :: Simplex A : Role A*.
@@ -19,11 +25,11 @@ Evidence base: repo README and code (`leanbench/`, 4,345 lines incl. tests), pro
| Lean | Simplex view |
|---|---|
| `(G : Type*) [Group G]` | role `G`, `context` = "carrier", `choices` = admissible type/class constraints |
| `(G : Type*) [Group G]` | role `G`, `context` = "carrier", type/class constraints as attributes; `choices` only for explicit finite value sets |
| `(h : P x)` hypothesis | `Fact` role |
| conclusion after the final `:` | `Goal` role |
| `def` / `structure` being introduced | `own` role (what the thing *is*) |
| a lemma used inside a proof | nested simplex (edge in the complex) |
| a lemma used inside a proof | explicit nested schema reference (when recorded) |
`leanbench/typeindex.py` already extracts exactly this split: `input` (binders and hypotheses), `goal` (the conclusion), `own` (definitions). So the type index is a simplex index in all but name. The new work is to make it a **first-class schema that steers inference** instead of a passive index.