Actuary
← Z.ai GLM 5.2

Evidence block

Instruction following · cycle 45

z-ai/glm-5.2 · 2026-08-05 14:43:56 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
94a4f695cc32b47a0f5b84bd8025108c459a15c2fed273cc365c4849ea806df6
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
$0.003219

Checker layer

Panel members.

# Item Item id Outcome Score
1 instruction-following-015 26a44f13-e4a8-4273-9cfd-5adce0f3eec9 ok 1.00
2 instruction-following-045 462a27a2-2b9a-417e-b2c3-b20776fad9a0 ok 1.00
3 instruction-following-075 38ede11b-3d52-4905-a6a4-8ee6684146e4 truncated 0.00
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 Item text stays unpublished so nobody can tune against the panel — identity travels as the item id and its pool source id.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 StreamLake z-ai/glm-5.2 gen-1785941000-8RsaKeKCoyXDOyKjniqi stop 10945 52 518 $0.000462 44343a812ededc65de068b9a8af8b8b6f7b71a15974252642487826a4796c667 2026-08-05 14:43:31 UTC
2 openrouter StreamLake z-ai/glm-5.2 gen-1785941000-z0v7PNHkbeMUWEfT5qgv stop 7457 61 313 $0.0002818 fcc5ce047f7912486ef1a6fab4c7df803541e1770a5919c95683e21b48bb0d17 2026-08-05 14:43:27 UTC
3 openrouter 55 2048 $0.0018083 2026-08-05 14:43:56 UTC
4 openrouter StreamLake z-ai/glm-5.2 gen-1785941000-yrDlp5b31J6YzsrnCp81 stop 12799 50 496 $0.0004425 4632166d4c917a5d482efe2d72cf06c591e6ceff8762ec97802216318d26dcba 2026-08-05 14:43:33 UTC
5 openrouter StreamLake z-ai/glm-5.2 gen-1785941015-xFTF39ZkzDpjOLb8ndgf stop 6049 61 247 $0.0002244 1ae4323e07007b052bb7f38ff4ab1ca0ef9e12498d537111ad88e9441ccde736 2026-08-05 14:43:41 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/94a4f695cc32…

curl -s https://actuary.sh/api/v1/evidence/94a4f695cc32b47a0f5b84bd8025108c459a15c2fed273cc365c4849ea806df6 \
  | 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.