Dataset Viewer
Auto-converted to Parquet Duplicate
file
stringlengths
13
31
sha256
stringlengths
64
64
claim_id
stringlengths
8
8
⌀
key_measurements
stringlengths
22
456
bench_sglang.json
4d2d3141150599a4a0af76336a261889ce1d99ec9a3fc12c1eb98c11ade0e035
null
{"done": true, "ts": 1785081682.6606686}
bench_vllm.json
df1b2f84d6f16e837cea464d4a85f8a440a1cd4e192e7b100521d88f73c0b423
null
{"done": true, "ts": 1785078842.62575}
declarative_gate_result.json
bd973a6900538cae27d2c9efd0af3cdac68fad3505970b2685b7945b16559ccf
claim_01
{"overall_pass": true}
frommem_EMPTY.json
24c3e6f1658d7283a0fe89102610398f2d5dfcdbc0f283c87cbfbfa39319fd49
null
{"memory_entries": 0, "total_correct": 0, "total_shifts": 0}
frommem_gemma4.json
8e92b63502e5f719cdfb4c82635c7ceed1d6f005c1a676c7d667bc597fbb70dd
claim_11
{"accuracy_pct": 100.0, "decode_tokens_total": 0, "harness_contains_solvers": false, "memory_entries": 9, "n_families": 9, "total_correct": 180, "total_shifts": 180, "wh": 6.541685964929519}
galahad_docgraft.json
62ee26ea26ac983ad66385e952ec62fc1c5249aaef3c4f32605b862cd5f8f9c1
null
{"ts": 1785077585.6192904}
galahad_qwen_4k.json
c99326bed0a46e33413176ee657e148efbc49dd74bdf98baa2c3d01bbbe63725
null
{"ts": 1785084489.0798562}
lean_geo_gemma4.json
9c450efe700479731e78736c745397c77b658e7b022ab165e4d2681eb1b1092b
null
{"deposit_tokens": 53, "deposit_tries": 1, "graft_reused_reverifies_0tok": true, "kernel_verified": true, "negative_control_rejected_wrong": true, "reuse_decode_tokens": 0}
logic_e2e_gemma4.json
baff3d7dec0de14b29759ba2441b7fb915995331ff24e46bb884efdec2359df4
null
{"deposit_tokens": 174, "deposit_tries": 1, "gate_accepted": true, "graft_reused_reverifies_0tok": true, "model_emitted_rules": true}
logic_meta_gemma4.json
a03d98544db8336d331de913f482119dd84146ab8ec66abc2be16f0cbeb44194
null
{"graft_reused_reverifies_0tok": true, "reuse_decode_tokens": 0, "sound_framework_accepted": true, "unsound_framework_rejected": true}
rerun_basediv_120fix.json
44f29c28d8fe220ef0e752ab7d1e1f252831a197d83e4b70df0a3c129b314d83
claim_14
{"cold_shift_correct": 9, "deposit_tokens_total": 899, "families_deposited": 1, "flywheel_shift_correct": 10, "flywheel_shift_decode_tokens_total": 0, "m_shifts_per_family": 12, "n_families": 1, "n_shift_evals": 12, "trivial_shift_cold_solves_both": 9, "true_transfer_cell": 2, "wh": 16.43622739100736}
rerun_basediv_multiinst.json
fc755610c17724bdbb4f8c68c80a8e854d87ef311c78d37a2deb9e9b2b486043
claim_15
{"cold_shift_correct": 9, "deposit_tokens_total": 1373, "families_deposited": 1, "flywheel_shift_correct": 12, "flywheel_shift_decode_tokens_total": 0, "m_shifts_per_family": 12, "n_families": 1, "n_shift_evals": 12, "trivial_shift_cold_solves_both": 9, "true_transfer_cell": 3, "wh": 17.619349718258828}
reuse_deepseek.json
134c57373d7a4eb15ae685667f5c6f575de94138bd405bbb6bb4f3a3acf26dc4
claim_16
{"accuracy_pct": 100.0, "decode_tokens_total": 0, "n_families": 9, "total_correct": 180, "total_shifts": 180, "wh": 6.321570693158303}
reuse_fastpath_proof.json
c397cc99dfdf18e4b240f18ea4e2813cd2b067e3266256c87390398ea89135b2
claim_17
{"adversarial_cases": 5, "adversarial_leaked": 0, "all_pass": true, "equivalence_mismatches": 0, "equivalent": true, "fastpath_correct": 140, "fastpath_ms_total": 1092.7, "fastpath_us_per_call": 7805.2, "ground_truth_total": 140, "not_breakable": true, "sandbox_correct": 140, "sandbox_ms_total": 102643.2, "sandbox_us_p...
reuse_phi4.json
ae611cd1c2e89dc080d05536995ffcca8fed4570cfa3d1643750e59e9ee9b1cf
claim_18
{"accuracy_pct": 100.0, "decode_tokens_total": 0, "n_families": 9, "total_correct": 180, "total_shifts": 180, "wh": 6.350301879550242}
reuse_qwen3.json
757256d7b7110c81f0e00210ff2532e76b33aab7fe67213219575260f737af6c
claim_19
{"accuracy_pct": 100.0, "decode_tokens_total": 0, "n_families": 9, "total_correct": 180, "total_shifts": 180, "wh": 6.4436211629438365}
reuse_speed_bench.json
aeda86dc2e2711f1e7e027021ef549f756d19865ce42d28de65442f81abd8501
claim_20
{"compile_once_speedup_x": 1.41, "compiled_cached_avg_us": 384.8408, "deposit_subprocess_spawn_ms": 38.42, "exec_every_call_avg_us": 541.4079, "iters_per_solver": 20000, "reuse_vs_sandbox_speedup_x": 100.0, "solvers": 9}
routing_bench.json
6451424344245be52f5c2e46de51a61825b67b34429cfc0b08fb2645392e1047
claim_21
{"index_entries": 9, "queries": 9, "routed_correct": 9, "under_5ms": true}
scaffold_e2e_gemma4.json
abed67c1573c06db300f402b2a98c1893a7ee96e3695cd30ab1bf6948abccd62
null
{"adaptation_accepted": true, "deposit_tokens": 187, "model_adapted": true, "tries": 1}
scaffold_gemma4.json
b0ad9afadb8998075042dd548757ea975340c9f0d322db58c5956aae8a617939
null
{"bad_adaptation_rejected": true, "graft_reuse_decode_tokens": 0, "mode3_adaptation_to_new_case_consistent": true, "scaffold_verified": true}
transfer_scale_v2_FINAL120.json
95f3f1b7541a9815bcab9da5d55b942f7f24bbcbf50e63612e6793ac829e22da
claim_24
{"cold_shift_correct": 0, "deposit_tokens_total": 16579, "families_deposited": 8, "flywheel_shift_correct": 150, "flywheel_shift_decode_tokens_total": 0, "m_shifts_per_family": 30, "n_families": 9, "n_shift_evals": 150, "trivial_shift_cold_solves_both": 0, "true_transfer_cell": 0, "wh": 81.12983232203962}
v2_strpal_external.json
da91ec0497d7c902942d1f547e652e61e8d4f8fec678bdff376c88ac8df12557
claim_25
{"decode_tokens_total": 0, "external_solver_agnostic_verified": true, "flywheel_accuracy_pct": 100.0, "flywheel_correct": 30, "n_shifts": 30}
verification_ladder.json
a26044e2f56c4bcfc9d30beb746141c603cefdf1fc4d692230e74940252eadf3
claim_26
{"buggy_caught_at_stage_by_ladder": true, "buggy_passes_naive_stage_only": true, "buggy_solver_accepted": false, "correct_solver_accepted": true, "n_fuzz": 150}
w_window_15M.json
22f03ef299c797b04bcda396e34c7f27fb3c631e27a055904ac57daebba2fdd5
claim_27
{"all_retrieved": false, "block_tokens": 32000, "blocks_deposited": 468, "deposit_s": 5545.088841199875, "peak_vram_mb": 24429, "store_tokens": 14976000, "target_tokens": 15000000}
window6m.json
536a5769b40415dccc56ab32212071190c03ec8d0c35dc39bba79b84f498a670
claim_28
{"access_s_max": 0.5931730270385742, "access_s_min": 0.551487922668457, "all_retrieved": true, "base_vram_mb": 24433, "block_tokens": 30000, "blocks_deposited": 200, "deposit_s": 3505.8332929611206, "peak_vram_mb": 24696, "store_tokens": 6000000, "target_tokens": 6000000, "vram_growth_mb": 263}

Galahad — SHA-256 Proof

Timestamped, independently verifiable measured benchmark results behind the Galahad paper.

📄 Paper: https://huggingface.co/papers/2607.23806 📚 arXiv: https://arxiv.org/abs/2607.23806 🧪 Live testbench (Space): https://huggingface.co/spaces/Corbenic/galahad-bench 💾 Proof mirror (GitHub): https://github.com/corbenicai/galahad-bench/tree/main/proof

What this is

This dataset is the integrity proof for the paper's claims: the raw measured outputs (pure numbers — token counts, latency in ms, VRAM in MB, accuracy, pass/fail) plus a full SHA-256 manifest. It contains no source code, no engine internals, and no solver contents — only measurements.

  • results/ — the raw measured benchmark JSON files.
  • index.jsonl — one row per result file (filename, its SHA-256, key numeric measurements) so the dataset viewer renders a quick overview.
  • results.sha256 — the hashes in sha256sum -c format.
  • SHA256_PROOF.txt — the full manifest: verifiable hashes plus sealed commitments to files kept private (source, verified-solver memory, internal notes, engine binaries). A sealed hash proves a file existed and was fixed on the stated date without revealing its contents.

Verify it yourself

git clone https://huggingface.co/datasets/Corbenic/galahad-bench-proof
cd galahad-bench-proof
sha256sum -c results.sha256          # expect: every line "OK"

Highlights (measured, single GPU)

  • Movable window — results/window6m.json: a 6,000,000-token context on one GPU at flat VRAM (base ≈ peak MB), correct retrieval at every depth, passing negative control.
  • Head-to-head — results/bench_vllm.json, results/bench_sglang.json: same GPU / model class; vLLM hard-stops at ~30k tokens, SGLang truncates past ~32k, Galahad runs to millions.
  • Verified reuse — results/reuse_*.json, results/frommem_*.json: 0-decode-token reuse across four model families; frommem_EMPTY.json is the negative control (empty memory solves nothing).

Scope & honesty

These are measurements, not the method. The engine, the verified-solver memory, and the internal method notes are intentionally not published — their SHA-256 commitments are in SHA256_PROOF.txt so the record is complete and tamper-evident. No trade-secret content is exposed by a hash.

load_dataset("Corbenic/galahad-bench-proof") loads the index.jsonl overview; the full result files live under results/.

Cite

arxiv.org/abs/2607.23806
Downloads last month
-

Paper for Corbenic/galahad-bench-proof