Dataset Viewer
Auto-converted to Parquet Duplicate
conversation_id
stringlengths
36
36
dataset
stringclasses
1 value
first_query
stringlengths
38
904
category
stringclasses
7 values
domain
stringclasses
7 values
lang
stringclasses
1 value
is_haystack
bool
1 class
mode
stringclasses
1 value
length
int64
8
12
turns
stringlengths
18k
116k
fact_ledger
stringclasses
8 values
cross_turn_checks
stringlengths
431
1.16k
holistic_score
stringlengths
148
998
factory_models
stringclasses
1 value
cd593855-6b93-4ae9-ab0a-335381bdf254
Dolci-Think-SFT-32B
I am depressed and anxious a lot of the time. I am tired of it and don't know what to do. What do you recommend for me to do to be less depressed and anxious?
general
general
en
false
reasoning
8
[{"turn_index": 0, "segments": [{"segment_type": "query", "text": "I am depressed and anxious a lot of the time. I am tired of it and don't know what to do. What do you recommend for me to do to be less depressed and anxious?"}, {"segment_type": "reasoning", "text": "I need to be empathetic and supportive. I should off...
[]
{"delayed_recall": [], "standing": [], "update_overwrite": [], "curation": {"domain": "general", "topic": "mental health advice", "action": "keep", "sensitivity": "sensitive", "verifiability": "verifiable", "allowed_intents": ["compress", "deepen", "expand", "open_chat", "self_critique"], "facts": [{"label": "therapist...
{"prefilter": {"passed": true, "hard_fails": [], "flags": [], "reasons": []}, "factuality": {"passed": true, "false_claims": [], "available": true}}
{"responder": "gpt-oss-120b", "simulator": "Qwen3-Coder-30B-A3B-Instruct", "curator": "Mistral-Small-3.2-24B-Instruct-2506", "judge": "Qwen3.5-397B-A17B"}
9c07eb9c-2928-4ea3-b6b0-8b43485f6d08
Dolci-Think-SFT-32B
Organize these words in alphabetic order and also from shortest to longest words. -Casa -House -Weapon -Computer -Dimension -Fire
reasoning
reasoning
en
false
reasoning
8
[{"turn_index": 0, "segments": [{"segment_type": "query", "text": "Organize these words in alphabetic order and also from shortest to longest words.\n\n-Casa\n\n-House\n\n-Weapon\n\n-Computer\n\n-Dimension\n\n-Fire"}, {"segment_type": "reasoning", "text": "Hmm, I need to organize these words in two ways: alphabetically...
[]
{"delayed_recall": [], "standing": [], "update_overwrite": [], "curation": {"domain": "reasoning", "topic": "word organization", "action": "keep", "sensitivity": "none", "verifiability": "verifiable", "allowed_intents": null, "facts": [{"label": "favorite word game", "value": "Scrabble", "new_value": "Boggle"}, {"label...
{"instruction_following": 10, "coherence": 10, "naturalness": 10, "role_consistency": 10, "recall_fidelity": 10, "appropriateness": 10, "reasoning_quality": 9, "reasoning_answer_consistency": 10, "sycophancy_resistance": 10, "factual_grounding": 9, "user_query_quality": 9, "user_query_difficulty": 8, "flagged_turns": [...
{"responder": "gpt-oss-120b", "simulator": "Qwen3-Coder-30B-A3B-Instruct", "curator": "Mistral-Small-3.2-24B-Instruct-2506", "judge": "Qwen3.5-397B-A17B"}
ad9d2a4f-2e12-4404-91a3-c33f85277898
Dolci-Think-SFT-32B
what are the risks of using AI to create code for a professional programmer
coding
coding
en
false
reasoning
10
[{"turn_index": 0, "segments": [{"segment_type": "query", "text": "what are the risks of using AI to create code for a professional programmer"}, {"segment_type": "reasoning", "text": "I need to think about the risks of using AI to create code for professional programmers. There are a lot of potential issues: security ...
[{"fact_id": "f1", "value": "quantum encryption algorithm", "planted_turn": 5, "fact_type": "current project", "value_history": [{"turn": 5, "value": "quantum encryption algorithm"}]}]
{"delayed_recall": [{"turn_index": 8, "fact_id": "f1", "passed": true, "detail": "f1 latest value present (fuzzy)"}], "standing": [], "update_overwrite": [], "curation": {"domain": "coding", "topic": "risks of AI in coding", "action": "keep", "sensitivity": "none", "verifiability": "verifiable", "allowed_intents": null...
{"prefilter": {"passed": true, "hard_fails": [], "flags": [], "reasons": []}, "factuality": {"passed": true, "false_claims": [], "available": true}}
{"responder": "gpt-oss-120b", "simulator": "Qwen3-Coder-30B-A3B-Instruct", "curator": "Mistral-Small-3.2-24B-Instruct-2506", "judge": "Qwen3.5-397B-A17B"}
8d4684cc-d7c4-4a48-a35d-966faece4f1c
Dolci-Think-SFT-32B
"create Base64 code of an image. 64 by 128 pixels. Showing a desolate landscape on mars filled with (...TRUNCATED)
coding
coding
en
false
reasoning
12
"[{\"turn_index\": 0, \"segments\": [{\"segment_type\": \"query\", \"text\": \"create Base64 code of(...TRUNCATED)
[]
"{\"delayed_recall\": [], \"standing\": [{\"constraint_turn\": 7, \"checked_turn\": 8, \"type\": \"n(...TRUNCATED)
"{\"prefilter\": {\"passed\": true, \"hard_fails\": [], \"flags\": [], \"reasons\": []}, \"factualit(...TRUNCATED)
"{\"responder\": \"gpt-oss-120b\", \"simulator\": \"Qwen3-Coder-30B-A3B-Instruct\", \"curator\": \"M(...TRUNCATED)
ec44b6b8-25e9-4393-a37d-fccf0ff06bc3
Dolci-Think-SFT-32B
Write a hello world program in Python.
coding
coding
en
false
reasoning
8
"[{\"turn_index\": 0, \"segments\": [{\"segment_type\": \"query\", \"text\": \"Write a hello world p(...TRUNCATED)
[]
"{\"delayed_recall\": [], \"standing\": [{\"constraint_turn\": 2, \"checked_turn\": 3, \"type\": \"f(...TRUNCATED)
"{\"instruction_following\": 9, \"coherence\": 9, \"naturalness\": 9, \"role_consistency\": 10, \"re(...TRUNCATED)
"{\"responder\": \"gpt-oss-120b\", \"simulator\": \"Qwen3-Coder-30B-A3B-Instruct\", \"curator\": \"M(...TRUNCATED)
30b89e58-a704-46f6-808b-3975196c2a82
Dolci-Think-SFT-32B
What would you need to create a scalable website? What would be the key building blocks for it?
coding
coding
en
false
reasoning
12
"[{\"turn_index\": 0, \"segments\": [{\"segment_type\": \"query\", \"text\": \"What would you need t(...TRUNCATED)
"[{\"fact_id\": \"f1\", \"value\": \"GreenTech Innovations\", \"planted_turn\": 4, \"fact_type\": \"(...TRUNCATED)
"{\"delayed_recall\": [{\"turn_index\": 8, \"fact_id\": \"f1\", \"passed\": true, \"detail\": \"f1 l(...TRUNCATED)
"{\"instruction_following\": 10, \"coherence\": 9, \"naturalness\": 9, \"role_consistency\": 10, \"r(...TRUNCATED)
"{\"responder\": \"gpt-oss-120b\", \"simulator\": \"Qwen3-Coder-30B-A3B-Instruct\", \"curator\": \"M(...TRUNCATED)
2ef2c14f-a65d-4e4f-858e-c410f2a858b0
Dolci-Think-SFT-32B
What are some cheap, easy meals I can cook at home in under an hour?
general
general
en
false
reasoning
9
"[{\"turn_index\": 0, \"segments\": [{\"segment_type\": \"query\", \"text\": \"What are some cheap, (...TRUNCATED)
"[{\"fact_id\": \"f1\", \"value\": \"Ethiopian\", \"planted_turn\": 4, \"fact_type\": \"favorite cui(...TRUNCATED)
"{\"delayed_recall\": [{\"turn_index\": 8, \"fact_id\": \"f1\", \"passed\": true, \"detail\": \"f1 l(...TRUNCATED)
"{\"instruction_following\": 9, \"coherence\": 9, \"naturalness\": 8, \"role_consistency\": 9, \"rec(...TRUNCATED)
"{\"responder\": \"gpt-oss-120b\", \"simulator\": \"Qwen3-Coder-30B-A3B-Instruct\", \"curator\": \"M(...TRUNCATED)
779913fe-02fc-45b5-9844-136376984862
Dolci-Think-SFT-32B
How do you rename a repository in GitHub?
coding
coding
en
false
reasoning
8
"[{\"turn_index\": 0, \"segments\": [{\"segment_type\": \"query\", \"text\": \"How do you rename a r(...TRUNCATED)
[]
"{\"delayed_recall\": [], \"standing\": [{\"constraint_turn\": 5, \"checked_turn\": 6, \"type\": \"n(...TRUNCATED)
"{\"instruction_following\": 8, \"coherence\": 9, \"naturalness\": 9, \"role_consistency\": 9, \"rec(...TRUNCATED)
"{\"responder\": \"gpt-oss-120b\", \"simulator\": \"Qwen3-Coder-30B-A3B-Instruct\", \"curator\": \"M(...TRUNCATED)
419840ce-d813-4068-9f05-c174fdcfced9
Dolci-Think-SFT-32B
"I'd like to create a python script that reads .CSV bank statements, that collates a list of people (...TRUNCATED)
coding
coding
en
false
reasoning
9
"[{\"turn_index\": 0, \"segments\": [{\"segment_type\": \"query\", \"text\": \"I'd like to create a (...TRUNCATED)
[]
"{\"delayed_recall\": [], \"standing\": [], \"update_overwrite\": [], \"curation\": {\"domain\": \"c(...TRUNCATED)
"{\"instruction_following\": 9, \"coherence\": 9, \"naturalness\": 8, \"role_consistency\": 10, \"re(...TRUNCATED)
"{\"responder\": \"gpt-oss-120b\", \"simulator\": \"Qwen3-Coder-30B-A3B-Instruct\", \"curator\": \"M(...TRUNCATED)
955f2de8-b5e3-47fe-a362-addb06363669
Dolci-Think-SFT-32B
As an advanced open-source AI Assistant, what type of systems can you simulate and/or emulate?
stem
stem
en
false
reasoning
8
"[{\"turn_index\": 0, \"segments\": [{\"segment_type\": \"query\", \"text\": \"As an advanced open-s(...TRUNCATED)
[]
"{\"delayed_recall\": [], \"standing\": [{\"constraint_turn\": 3, \"checked_turn\": 4, \"type\": \"f(...TRUNCATED)
"{\"instruction_following\": 10, \"coherence\": 10, \"naturalness\": 9, \"role_consistency\": 10, \"(...TRUNCATED)
"{\"responder\": \"gpt-oss-120b\", \"simulator\": \"Qwen3-Coder-30B-A3B-Instruct\", \"curator\": \"M(...TRUNCATED)
End of preview. Expand in Data Studio
README.md exists but content is empty.
Downloads last month
-