Actuary
← Z.ai GLM 5.2

Evidence block

Instruction following · cycle 21

z-ai/glm-5.2 · 2026-08-04 13:45:00 UTC

This page is the full public record of one completed panel — the record’s hash-committed unit of evidence. Its identity is its hash; nothing here can change without changing the URL.

Identity

What this block is.

Panel mean
0.800
Members
5
Checker scored
4
Frozen floor
0.700
Anytime level α
0.050
Hashes and versions
Evidence hash
7a747b33a92ab23a6e5620d51aa5eb3a46ea452c675d7e3fc630053ef51c875b
Design hash
4f5cc6b9ff4ce026ccab93f5a3c8241f78eb005bc8b8068e4c701e9236504412
Panel hash
e7dfea1f343f7bf37e066a1117ca21f3438e804ad61437bf72d810640cef7e71
Stream
1ca75a9f-6952-4c04-9b8a-1049aaa718a2
Epoch
37cb9150-9ea9-4f2f-baa9-815ec952f905
Pool version
instruction_following-v3
Checker version
instruction_following-v1
Panel size
5
Provider spend
$0.009159

Checker layer

Panel members.

# Item Item id Outcome Score
1 instruction-following-015 26a44f13-e4a8-4273-9cfd-5adce0f3eec9 rate limited 0.00
2 instruction-following-045 462a27a2-2b9a-417e-b2c3-b20776fad9a0 ok 1.00
3 instruction-following-075 38ede11b-3d52-4905-a6a4-8ee6684146e4 ok 1.00
4 instruction-following-105 9294d9d8-5c71-4d76-b053-b5cefc08d0e1 ok 1.00
5 instruction-following-135 7d213d46-039d-4f3a-aecf-c785ef61cc80 ok 1.00

Deterministic checkers score every response in [0, 1]; a member that ends in a terminal failure scores 0 and stays on the record. Item text stays unpublished so nobody can tune against the panel — identity travels as the item id and its pool source id.

Capture layer

Provider requests.

# Provider Upstream Response model Response id Finish Latency ms Tokens in Tokens out Cost Output SHA-256 Observed
1 openrouter $0.0042087 2026-08-04 13:44:42 UTC
2 openrouter Decart z-ai/glm-5.2 gen-1785851082-0f7HTNwjOQ0a9S7OlMLU stop 2156 61 288 $0.0004441 fcc5ce047f7912486ef1a6fab4c7df803541e1770a5919c95683e21b48bb0d17 2026-08-04 13:44:45 UTC
3 openrouter Decart z-ai/glm-5.2 gen-1785851085-fxRxQWd7Tp0ybKea1igJ stop 15221 55 1996 $0.003005 3792adf5f1c853573c2331a89c29320dcc87acef2527ac99e200b0cfe58acceb 2026-08-04 13:45:00 UTC
4 openrouter Decart z-ai/glm-5.2 gen-1785851084-r2cPturZ0rq7RCt1ddd7 stop 3239 50 319 $0.0004885 4632166d4c917a5d482efe2d72cf06c591e6ceff8762ec97802216318d26dcba 2026-08-04 13:44:48 UTC
5 openrouter Decart z-ai/glm-5.2 gen-1785851091-0BH40OUAWvA07KjdvEJr stop 5996 61 667 $0.0010126 1ae4323e07007b052bb7f38ff4ab1ca0ef9e12498d537111ad88e9441ccde736 2026-08-04 13:44:57 UTC

Everything a probe captures, as captured. Raw model text is scrubbed thirty days after the run; its SHA-256, the outcome and the score are permanent. Costs are provider-reported and include reserved estimates where a run’s true cost was unrecoverable; a dash is a field the provider did not return.

Verify

Recompute the hash yourself.

The JSON twin of this page carries the exact evidence object this URL is named after. Canonicalise it per RFC 8785 and hash it; the digest must equal the address of this page.

GET /api/v1/evidence/7a747b33a92a…

curl -s https://actuary.sh/api/v1/evidence/7a747b33a92ab23a6e5620d51aa5eb3a46ea452c675d7e3fc630053ef51c875b \
  | jq -cjS .evidence | shasum -a 256

For this object’s field types, jq -cjS reproduces RFC 8785 canonical form, and the JSON twin serves the hashed bytes verbatim. Blocks of exactly this shape appear verbatim in signed certificates — verify one offline, or read how the e-process turns panel means into an alarm.