Datasets:
schema_version string | example_id string | role string | variant string | split string | reviewer_id string | target_id string | trajectory_id string | turn_index int64 | criterion_index int64 | candidate_index int64 | loss_mask_type string | messages list | metadata_json string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
amazon-c2-direct-distillation-sft-row-v1 | 9d7f8d1140df53a54c6f123ef08a9c8024325b67a3f5e6ee239a7d5e80facb78 | rubric_judge | non-diverse | train | amzd_448195997decb7183b61cbce | amzt_e9c2fdea41a03000178ca546 | 9e4445cae149f733d8a830d4e8677bbf49d7eb20a697abe90777c062cd1a37e6 | 0 | null | null | qwen3_final_response | [
{
"role": "user",
"content": "Score EACH candidate Amazon product review independently against the weighted rubric for what this reviewer would likely write about this product. Do not reward general review quality.\n\n<|The Start of Rubric|>\n<rubric>\n### Enthusiastic positive overall verdict\n**Criterion:... | {"candidate_count":40,"candidate_order":[1,38,36,11,7,35,23,10,32,24,2,0,8,33,19,18,13,37,21,30,27,22,16,39,9,12,34,29,3,20,17,6,26,28,15,14,31,25,5,4],"criterion_count":1,"rubric_sha256":"10bb0d6ab8e81ab4e85baacbda5f279ebb77b800e13268477e868d15b3599e5e","scores_sha256":"381d3a98eef67045b1c734703cdeccd17c23afe7fc4966f7... |
amazon-c2-direct-distillation-sft-row-v1 | 4964b59730273a898fe4ba250356238a53a19747dc312c9d08c15eea07e9c0a7 | rubric_judge | non-diverse | train | amzd_f5fb875941d8aab4ad8e0008 | amzt_9d633974a88c4f8703b388d7 | 902599a9f68d342b06ab9b02c84e5c8c9c7ab0da88e13364e50f356b9e4a9407 | 0 | null | null | qwen3_final_response | [
{
"role": "user",
"content": "Score EACH candidate Amazon product review independently against the weighted rubric for what this reviewer would likely write about this product. Do not reward general review quality.\n\n<|The Start of Rubric|>\n<rubric>\n### Overall positive verdict\n**Criterion:** The review... | {"candidate_count":40,"candidate_order":[0,14,1,39,28,5,3,37,8,7,23,11,20,19,18,29,27,33,4,15,30,26,21,12,9,16,34,10,2,35,32,13,31,25,17,36,22,38,6,24],"criterion_count":16,"rubric_sha256":"40dcc5f83ed4b576c54f1f894b014fed8e12392af51a3e333970cf20ef761154","scores_sha256":"6c94651f5aaced8a6645ae98f30a4faedff4ad4279ef6d7... |
amazon-c2-direct-distillation-sft-row-v1 | ec4a18f490284cd01278eaa5b74b77ded24f13ba7c30bb77cd35ec51a5a4e61f | rubric_judge | non-diverse | train | amzd_f103f12060a2c3e2ce97a81e | amzt_4f9dd49d60156e4e39e9f86f | 5efcbf9b0d2bb73a9a0d27040bb073876cf403c2ac2c78d7152318b1ded9f017 | 0 | null | null | qwen3_final_response | [
{
"role": "user",
"content": "Score EACH candidate Amazon product review independently against the weighted rubric for what this reviewer would likely write about this product. Do not reward general review quality.\n\n<|The Start of Rubric|>\n<rubric>\n### Judges mop durability against an expected lifespan\... | {"candidate_count":40,"candidate_order":[16,25,33,22,27,13,6,14,3,8,38,7,29,39,24,31,12,28,2,18,36,20,5,23,9,0,35,34,26,32,1,4,30,19,37,21,15,10,11,17],"criterion_count":3,"rubric_sha256":"2842dbb5e3e90ab947a8691456b6286b25a2ee031ba2d0135f0eb537420333b6","scores_sha256":"2c2424fa735f938a36178087169e6ace358d8fe28826d084... |
amazon-c2-direct-distillation-sft-row-v1 | 3fd932baf15a961d9ce5f63d58caa2f28c11ba4d6f527d3d1e24995ae78d7796 | rubric_writer | non-diverse | train | amzd_22274c9ab58b7908d8bca656 | amzt_ddd286755b1226add07569d7 | 20acec6f82491cfb826359cfcdaff76d6ff62189a996019ce203924310b8a3b7 | 0 | null | null | qwen3_final_response | [
{
"role": "user",
"content": "You are an expert at characterising how one specific person writes Amazon product reviews.\n\nYou will be shown 8 reviews written by ONE reviewer, each preceded by the corresponding product metadata. Using only these reviews, write a rubric that captures this reviewer's distinc... | {"candidate_count":40,"criterion_count":1,"rubric_sha256":"01facd3b08d88b5736971300c313c4faae591cb845307385269c8eeeea8bccae","source_example_ids_sha256":"378ce9913bc26f3362e8d4024e15550e83ad484428218234d43a2ad8d56a9875","source_oracle_turn":1,"source_trajectory_id":"20acec6f82491cfb826359cfcdaff76d6ff62189a996019ce2039... |
amazon-c2-direct-distillation-sft-row-v1 | c3990ebb3f0924688f9b3cea601aed725f66e73b9a2bda87ebb645727065d9d7 | rubric_judge | non-diverse | train | amzd_2b47cd98f5f9b19421d71f60 | amzt_133b27f21234d045da3fd9c7 | 19bceeaa0c94b6196aca9e0bdb8902c9306e569781fed4e1cd413cf474cad60d | 0 | null | null | qwen3_final_response | [
{
"role": "user",
"content": "Score EACH candidate Amazon product review independently against the weighted rubric for what this reviewer would likely write about this product. Do not reward general review quality.\n\n<|The Start of Rubric|>\n<rubric>\n### Big-guy build and gi fit framing\n**Criterion:** Ce... | {"candidate_count":40,"candidate_order":[15,1,5,23,39,3,26,18,34,10,37,33,31,24,20,11,32,29,9,36,35,21,38,16,30,22,4,2,14,12,13,6,0,8,19,25,17,7,27,28],"criterion_count":1,"rubric_sha256":"cb56c26408c02a4e5b098744bd1162ff89a4a0639eb5543cb8c606fc6098fe6e","scores_sha256":"d4f06b90554dc71bbe25c92f62df5813410038f732479822... |
amazon-c2-direct-distillation-sft-row-v1 | 178df321738df4bd75f92ff86a97eb23a682482379a3d8c75b93c0d952ac3320 | rubric_writer | non-diverse | train | amzd_b25b659aae1bf7a3830b86a2 | amzt_a48bfd72e7d7a0c3d6145b71 | 3eb94be6d73b7c1baf3cc22b05a17bc4b8c145918e8c81266410a7c15a41202d | 0 | null | null | qwen3_final_response | [
{
"role": "user",
"content": "You are an expert at characterising how one specific person writes Amazon product reviews.\n\nYou will be shown 8 reviews written by ONE reviewer, each preceded by the corresponding product metadata. Using only these reviews, write a rubric that captures this reviewer's distinc... | {"candidate_count":40,"criterion_count":1,"rubric_sha256":"0894b162859847bfb4dc405c9694df22dc208f7b322701b191eb435fe31f2c22","source_example_ids_sha256":"e90b2fbfbd1411041c6f2ad78da7bc54fbc40726414a26f99a87e06930d38200","source_oracle_turn":1,"source_trajectory_id":"3eb94be6d73b7c1baf3cc22b05a17bc4b8c145918e8c81266410a... |
amazon-c2-direct-distillation-sft-row-v1 | 022f2e46e5308026eb39a84275dd4b596d0d66b875b9e75919fb0ccb2184260e | rubric_writer | non-diverse | train | amzd_9c945afb912799ecce4a38a3 | amzt_862dec92f9fa07278df87472 | f2ae942ae25091d2721ac45b92f548cf8022c0edc609293597b8da12d17b3c62 | 0 | null | null | qwen3_final_response | [{"role":"user","content":"You are an expert at characterising how one specific person writes Amazon(...TRUNCATED) | "{\"candidate_count\":40,\"criterion_count\":4,\"rubric_sha256\":\"7200be24aef213b3be1df1836e0ee1faa(...TRUNCATED) |
amazon-c2-direct-distillation-sft-row-v1 | 0204f1f37cb1c42652cd7b84616abf81e4d75ce6b77f77528f8ffcfcb04dc397 | rubric_writer | non-diverse | train | amzd_3a59f636829d8990d60c3f4b | amzt_1e451b7390551776653b8058 | 7a6338110e07cf579e1afc715040ea4fe05afaf54aea5c4d5e743e923e26fc52 | 0 | null | null | qwen3_final_response | [{"role":"user","content":"You are an expert at characterising how one specific person writes Amazon(...TRUNCATED) | "{\"candidate_count\":40,\"criterion_count\":1,\"rubric_sha256\":\"b9842edce5fa091dd2f14e7c4ef1fd9e2(...TRUNCATED) |
amazon-c2-direct-distillation-sft-row-v1 | 89de0df27477f067e07fa4fc3a7470a8eb8ecafd85f5fca9f193fb9e50a72d7e | rubric_writer | non-diverse | train | amzd_fb8f48fd2f3f68302abef5ec | amzt_6d023f3a403cf146dcc90d61 | a26a781d13b9157f770ca0e2510512d438d650857ff4e7978ed385596659896e | 0 | null | null | qwen3_final_response | [{"role":"user","content":"You are an expert at characterising how one specific person writes Amazon(...TRUNCATED) | "{\"candidate_count\":40,\"criterion_count\":4,\"rubric_sha256\":\"d332c8c47f45af7eb95a13779f9bcda2b(...TRUNCATED) |
amazon-c2-direct-distillation-sft-row-v1 | c66777b08a854b891676dd523b4fe9159c1eca2fa4d03d73613baeb90b3bbed4 | rubric_judge | non-diverse | train | amzd_76dd2ac88560e2f87c68d9ba | amzt_8f5947be84e517995e0013bf | 50389a3e6e36171cfadd84f6f87ecbc49b8765a67fb114b19b704fd02f40b3f3 | 0 | null | null | qwen3_final_response | [{"role":"user","content":"Score EACH candidate Amazon product review independently against the weig(...TRUNCATED) | "{\"candidate_count\":40,\"candidate_order\":[10,6,12,22,17,16,7,4,2,20,5,9,24,38,26,37,32,27,13,33,(...TRUNCATED) |
End of preview. Expand in Data Studio
Amazon c2 quality-filtered distillation
Each retained trajectory contributes one direct rubric-writer target and one full-rubric listwise judge target over the original 40 candidates. Teacher scratch reasoning is discarded. Membership follows the pinned C11 quality-filter policy. Signed aggregate filter provenance is under quality/<split>/; it is not exposed as another dataset configuration.
The six configurations cross the two candidate variants with the three frozen stopping objectives. Every
configuration exposes only its combined training view and preserves the native train, validation, and test
splits.
| Config | Train | Validation | Test |
|---|---|---|---|
latent-state-spearman |
9,376 | 86 | 100 |
latent-state-best32-norm-tr |
8,598 | 98 | 96 |
latent-state-harmonic32-norm-tr |
9,130 | 104 | 102 |
non-diverse-spearman |
10,482 | 128 | 108 |
non-diverse-best32-norm-tr |
9,738 | 104 | 98 |
non-diverse-harmonic32-norm-tr |
10,478 | 116 | 112 |
Provenance
asingh15/amazon-c11-distillation-filteredatf3cbaf3082ca95ba94db218810b8b3e4fa076e81(membership-and-combined; catalog1b756f6f8b6124b7682db6a65c4054265b034b3e530eb03558fa675fad0e6f98)asingh15/amazon-c11-distillationat3f7302f2eb78cfa8a90110bd370237fd11e1638c(full-judge; catalogfaff332d2bf465d37c7d6e9c0bb068522d80e3ae1e5f777688ec6c607f99f1d2)- Derived catalog SHA-256:
99e4e166fb0baad71283cee2f60a26ecefb2e3291c41f7b361285dc6c1601d47 - Row schema:
amazon-c2-direct-distillation-sft-row-v1
- Downloads last month
- 274