relbench
Datasets
All datasets matching “relbench”relbench-v1
RelBench v1 databases
The original RelBench v1 relational databases and tasks, each
database in its own subdirectory in the self-describing manifest layout (plain parquet +
manifest.yaml):
<dataset>/
manifest.yaml # tables, primary keys, foreign-key graph, val/test timestamps
schema.svg # ER diagram
db/*.parquet # relational tables (plain parquet)
tasks/<task>/manifest.yaml # task spec (+ duckdb SQL for `forecast`… See the full description on the dataset page: https://huggingface.co/datasets/stanford-star/relbench-v1.relbench-v2-extra
RelBench v2 extras
Everything RelBench v2 added on top of v1: the v2-only
databases (with their tasks), and the v2-only tasks defined on the v1 databases. The v1
databases and tasks stay in
stanford-star/relbench.
Same self-describing manifest layout (plain parquet + manifest.yaml):
<dataset>/
manifest.yaml # tables, primary keys, foreign-key graph, val/test timestamps
schema.svg # ER diagram
db/*.parquet # relational… See the full description on the dataset page: https://huggingface.co/datasets/stanford-star/relbench-v2-extra.relbench-preprocessedrelbench-rawrelbench_O2
RelBench-O2: Constant-Time Assembly at -O2
Overview
This dataset collects the rel_bench constant-time programs, re-cut into one
self-contained program per record, compiled with LLVM/Clang-17 at -O2 for three
targets, and checked for constant-timeness with binsec -sse -checkct.
Constant-time source code is not enough: the optimizer can reintroduce a secret-dependent
branch or memory access that the -O0 build does not have, and it does so differently per
target. The… See the full description on the dataset page: https://huggingface.co/datasets/Akirayasha/relbench_O2.relbench-leaderboard