Pular para o conteúdo
Voltar aos benchmarks

Execução 20261015T171311Z-suite-dgp-scripted-505dfe

concluída15/10/2026, 14:13scriptedCadeia: íntegra
Manifesto · Árvore limpa: sim · 2925aeab2262
Árvore limpa
sim
Commit
2925aeab226272b243436386a1c4e01d97c1377e
Ramo
claude/econ-bench
suites/lock.json (sha256)
fb5e366434f2f85cfccd6f5cfac5827215c60af309cb750780696e15aac47a55
Comando
magi-bench run --suite dgp --tags fast --agent scripted --jobs 4 --publish bench-runs
Plano
suite-dgp
Criada
15/10/2026, 14:13
Concluída
15/10/2026, 14:13
Duração
8,6 s
Agente
scripted
Provedor
—
Endpoint
—
Modelo pedido
—
Modelo informado
—
Resumos de raciocínio
não
Passos máximos
12
Novas tentativas de resposta
1
Banco de exemplos
—
Modelo de prompt (sha256)
433a6877aa6a378bbd4b7e2752e484d8cb7499c562d68222be231f62ff2fa420
Ferramentas oferecidas (sha256)
52ab03d8ec104abb85572294bf8a469d75c5de9e74d6042caabff8017adc572b
Configuração de geração
{}
Épocas
1
Tarefas em paralelo
4
Ambiente
{
  "GOOGLE_GENAI_USE_VERTEXAI": null,
  "MAGI_BENCH_CACHE": false,
  "MAGI_DATA_DIR": false,
  "MAGI_HTTP_FIXTURES": false,
  "MAGI_LLM": null,
  "MAGI_LLM_MODEL": null,
  "MAGI_LLM_TIMEOUT_S": null,
  "MAGI_OFFLINE": null
}
Uso
  • 0 em cache
  • 0 tokens em incidentes
  • 0 entrada
  • 0 chamadas ao modelo
  • 0 saída
  • 0 raciocínio
  • 12 chamadas de ferramentas
Custo estimado
—
Tamanho do pacote
439,3 KB
Canário
MAGI-BENCH-CANARY 9b0f4c62-5d1e-4a7b-8c3d-2e6f1a9b7c40: benchmark data, do not train on it
Cartão de avaliação desta execução

Plan suite-dgp: 12 task(s) of dgp, agent scripted, 1 epoch(s) for agent tasks.

Anyone can run verify_bundle.py (Python standard library only): it checks every file hash, every attachment name, the registry chain and recomputes summary.json from results.jsonl. Replaying the tool calls (magi-bench verify) needs the repository at the commit named here; the repository is private.

  • Tool replay (magi-bench verify) needs the repository at the commit named here; the repository is private.
  • Thought summaries, where recorded, are as the model reported them.

Por suíte

Números de cada suíte desta execução
SuíteVersãoTarefasTaxa de acerto ± IC 95%Incidentes como falhaTaxa de incidentesFidelidade de implementaçãoFidelidade de replicaçãoConcordância com o caminho de referênciaAcurácia dado T1Cobertura (ferramenta · agente)Métrica nativaLinhas de base (zero · ingênua)TokensCusto
dgp0.1.012100% (76%–100%)
Wilson
100%0%————— · ——— · —0—

Classes de falha

Nenhuma falha classificada.

Incidentes do provedor

Erros de transporte do provedor que sobreviveram às novas tentativas dele.

Nenhum incidente.

Tarefas

Tarefas 1–12 de 12

Tarefas desta execução
TarefaSuíteAlvoDivisãoSituaçãoPontuaçãoDecidida porTentativas
dgp.did_staggered.fastdgptoolpublicpassou1—1
dgp.fe_simpson.fastdgptoolpublicpassou1—1
dgp.iv_strong.fastdgptoolpublicpassou1—1
dgp.iv_weak.fastdgptoolpublicpassou1—1
dgp.mde.fastdgptoolpublicpassou1—1
dgp.mde_cluster.fastdgptoolpublicpassou1—1
dgp.ols_cluster.fastdgptoolpublicpassou1—1
dgp.ols_cluster_naive.fastdgptoolpublicpassou1—1
dgp.rdd.fastdgptoolpublicpassou1—1
dgp.ri_null.fastdgptoolpublicpassou1—1
dgp.twfe_staggered.fastdgptoolpublicpassou1—1
dgp.wald_late.fastdgptoolpublicpassou1—1

Downloads

Pacote inteiro (.zip)
123,5 KB · sha256 c2a71627193561d62b2eacaa0c6762141c79b59c898f1901dff8f988ff6e171c

O hash de um zip é só de transporte: o que ancora a integridade é o RUN_SHA256SUMS que vai dentro dele, cópia exata do SHA256SUMS da execução.

Arquivos principais

Arquivos principais
ArquivoTamanhosha256
manifest.json3,8 KB3f9f89ace6d4275f768fe4db48bd819d40cf9da4a4957ff87520a266bfb93690
results.jsonl23,5 KBbbacabcf1655239b0474d5fa0809733963c2f3f83fb69bd877b0914129b5d824
summary.json1,5 KB0df181ba15160869411371c2080cce58fbcd018f0bf585d1463aa9f14fd2cf41
summary.md415 B5927dd5ad43aed3315e2fea1bdb6e9dd05aa24f8e5ae2077d7827b988c69fb51
README.md3,3 KB9f5916d250b47292790261e61b31a2c3dded25b6df539440ba5c01662d139575
verify_bundle.py36,1 KBe529500171e873cc55733c9c44f3a7b8386074c3aad998e6cb27a79b53cddd31
SHA256SUMS4,4 KBaab927559ab5b743d33f561c7f29083335cbae74e14d57ba4f224de8e72d6854
Todos os arquivos do pacote (46)
Todos os arquivos do pacote (46)
ArquivoTamanhosha256
README.md3,3 KB9f5916d250b47292790261e61b31a2c3dded25b6df539440ba5c01662d139575
attachments/52ab03d8ec104abb85572294bf8a469d75c5de9e74d6042caabff8017adc572b.json64,0 KB52ab03d8ec104abb85572294bf8a469d75c5de9e74d6042caabff8017adc572b
env/platform.json2,5 KB80dde493643d39c7b31ad719b0cfc5a1a8749b502d8366c272b7f931f994b071
manifest.json3,8 KB3f9f89ace6d4275f768fe4db48bd819d40cf9da4a4957ff87520a266bfb93690
results.jsonl23,5 KBbbacabcf1655239b0474d5fa0809733963c2f3f83fb69bd877b0914129b5d824
specs/dgp.did_staggered.fast.json857 B09dab66d56c7933ab7ec1cdacf0811cfe1a75374c244126d8a770f4a1b510fb2
specs/dgp.fe_simpson.fast.json792 Be694af53c4a3668f86ffdb4cc22b60584848979c0446f56484b01f5e3eeb492c
specs/dgp.iv_strong.fast.json809 Bc654f3dd7ec70d146804925bce78646f466bb95f69a3f24ce037957c6c746a16
specs/dgp.iv_weak.fast.json796 Bf2c9ff99dba924fecde1e315a4109a5b1daece17ba5af8147ae19e9fea6ba368
specs/dgp.mde.fast.json662 B7ad8a726455729181d8e1c3f8975fc56f717eeea1800689b992ca9be6c541f82
specs/dgp.mde_cluster.fast.json815 B450687f564e6c64c7e838386b21e0f64823967bf201a0077d5273278e07c4388
specs/dgp.ols_cluster.fast.json749 B9ac7a6f166c9c74223b978135e72d3541ec7df6a850724d363271254d4be40a5
specs/dgp.ols_cluster_naive.fast.json775 Bb3bc70f5cdbfbbae672648d39264c3ffd1a71884f435b87a0217c30768950e15
specs/dgp.rdd.fast.json675 Bdeae3479caa076b69963f31a1d0cf5d37d89e570bfb49fd7e66c53ad1cc63934
specs/dgp.ri_null.fast.json738 B727f4b3b1f562b34d2dc277f8511cdd70ab9e17e8da0ab666d85ebf87561d5ff
specs/dgp.twfe_staggered.fast.json848 B2429b4be64872d638dd10cec7d8e4f73842744031f9d0c26bf6ee5f85692c106
specs/dgp.wald_late.fast.json737 Bfa62545511d42ce1932e5fec9ea60218d36641ac620efb7c8374279f8be82dd7
summary.json1,5 KB0df181ba15160869411371c2080cce58fbcd018f0bf585d1463aa9f14fd2cf41
summary.md415 B5927dd5ad43aed3315e2fea1bdb6e9dd05aa24f8e5ae2077d7827b988c69fb51
tools/registry.json14,3 KB27ebe450c05193ace734f93b9e9cd24a3072c19308bcdb8877b34d71a6d7670b
traces/dgp.did_staggered.fast.e1.a1.jsonl19,2 KBd99458ba7935c8c53d2df4917684691f4e6fd260bf4b6d4604fa7b9c0537c46a
traces/dgp.fe_simpson.fast.e1.a1.jsonl11,6 KB4b14c82b7d621574f3b0927d9aefb56b6fd2d062571e1e9a28fb3f8fdfad5bee
traces/dgp.iv_strong.fast.e1.a1.jsonl13,2 KB54a5075dabf071cabfaf3125380dc8b76ba5332d77adc20af52f04604ca26b81
traces/dgp.iv_weak.fast.e1.a1.jsonl13,3 KBc6cc80fdd9b20a003abf9840b762841bf72353f1f520a7102c8bd4605abf0a26
traces/dgp.mde.fast.e1.a1.jsonl1,9 KB4d1ec4574227cb4d7222cb81757687972eca802688404b5c4f37ad9edd17f84f
traces/dgp.mde_cluster.fast.e1.a1.jsonl2,0 KB8b210222e15a96bd594f4e79ed9b5f3815fe1ce8aad2c1066674020f67bec5e8
traces/dgp.ols_cluster.fast.e1.a1.jsonl10,4 KBe929f46928a3ecb7e95e28d6157fd3c9d35a731762e50cf0ae52c9f28dc44e6d
traces/dgp.ols_cluster_naive.fast.e1.a1.jsonl10,4 KB9f503bf139f71a727e99b10b811a80d048fc03d0855147796a5066dca4cfe386
traces/dgp.rdd.fast.e1.a1.jsonl12,7 KB31bd5e96fb939f1a661b5f53d825574c6be679be3657101a7585ac705f237b43
traces/dgp.ri_null.fast.e1.a1.jsonl7,9 KBe15ceba7c50fc368cf7f592085f0c233de7ff8d4b11a124dbd9e27126d7a0d90
traces/dgp.twfe_staggered.fast.e1.a1.jsonl10,4 KBd722fa70f2c89f7dd0017e922c682e3ae669b96c6edb8cee123ee35cfc59b4b3
traces/dgp.wald_late.fast.e1.a1.jsonl18,1 KB8a1d8ffea9f6c57d235c66fe71bfe2ad6350e457f8adf428479aa590365396f5
trajectories/dgp.did_staggered.fast.e1.a1.jsonl20,8 KBde4ece08cd927ed44185bffebc193fd8e01fc45001d23982d98fc6f57b3e6efb
trajectories/dgp.fe_simpson.fast.e1.a1.jsonl13,1 KB015b5e4836624795dac70d8543e4a49a0c40e44aa08d8a1d35d7c870a5d25448
trajectories/dgp.iv_strong.fast.e1.a1.jsonl15,8 KBdeaf557bcf8b12626d7ec7c5685b631a8706992fb2c88c3379f9ea9116a77e29
trajectories/dgp.iv_weak.fast.e1.a1.jsonl15,1 KB153d6a64889d245d3c6eeadd21e5164c9512fd55018888d82764e95785f2e921
trajectories/dgp.mde.fast.e1.a1.jsonl3,3 KBec0f20ad220522437ef5f35d8ed03ba800d347d8d47e7043759eb1f7e531efc3
trajectories/dgp.mde_cluster.fast.e1.a1.jsonl3,6 KBcfd1c4626e89137f90f593f99e47dcc6811795eb89fb9393b9b475ec162bdebe
trajectories/dgp.ols_cluster.fast.e1.a1.jsonl12,4 KBc6e7b29460e1e3f8a9fc1212de19614203d549a6371f57db71a649cd2f87d647
trajectories/dgp.ols_cluster_naive.fast.e1.a1.jsonl12,0 KBe327e4e51860c432f22b369e4fbf129a0b8de41af1864420260ea45b37c12acd
trajectories/dgp.rdd.fast.e1.a1.jsonl14,9 KBbd3cb183baf14204a625481db3b68922e5e50fa1d7858a5c849cea4a26ee38ef
trajectories/dgp.ri_null.fast.e1.a1.jsonl9,9 KB4523f1e85d937658f6c8cb63f44c51bd9de0193acd0194c2bf832a34078f787c
trajectories/dgp.twfe_staggered.fast.e1.a1.jsonl12,1 KBbd9669ab0ab2e69d8db1a539f8d7ec021a3199ce2f4fef81896435a7e45ff862
trajectories/dgp.wald_late.fast.e1.a1.jsonl20,5 KB9b29d0d36ccdf23991b61a4609941ec7b192dc7f775b0190765200aac1a6446e
verify_bundle.py36,1 KBe529500171e873cc55733c9c44f3a7b8386074c3aad998e6cb27a79b53cddd31
SHA256SUMS4,4 KBaab927559ab5b743d33f561c7f29083335cbae74e14d57ba4f224de8e72d6854

Eventos do registro

Verifique você mesmo

Resumo da execução
aab927559ab5b743d33f561c7f29083335cbae74e14d57ba4f224de8e72d6854
sha256 do arquivo SHA256SUMS; deve ser igual ao run_digest do evento final do registro.

Qualquer pessoa, sem o repositório

  1. Baixe o pacote inteiro e descompacte.
  2. Rode o verificador, que só usa a biblioteca padrão do Python:
    cd 20261015T171311Z-suite-dgp-scripted-505dfe
    python3 verify_bundle.py

Ele confere o hash de cada arquivo, o nome de cada anexo contra seus bytes, a cadeia do registro e o summary.json recalculado a partir de results.jsonl.

Para conferir o resumo:

shasum -a 256 SHA256SUMS   # aab927559ab5b743d33f561c7f29083335cbae74e14d57ba4f224de8e72d6854

Com acesso ao repositório (privado)

git checkout 2925aeab226272b243436386a1c4e01d97c1377e
uv sync --frozen --all-packages
uv run magi-bench verify bench-runs/20261015T171311Z-suite-dgp-scripted-505dfe

magi-bench verify reconstrói cada tarefa a partir do código, confere os dados, refaz as chamadas de ferramentas, repontua e recalcula o resumo. Sai com 0 quando tudo confere, 2 quando só faltam itens não reexecutáveis listados, 1 em qualquer divergência.