messages
listlengths
3
3
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: Who is responsible for mind-mem maintenance?\n\nCandidates:\n[B-4567] David Park reported ...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are a memory compression expert. Given a set of conversation memory excerpts and a question, extract ONLY the facts relevant to answering the question.\n\nRules:\n1. Output a numbered list of factual observations derived from the context.\n2. Each observation must be a si...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are extracting evidence from conversation excerpts for an adversarial question.\nYour job is to extract relevant evidence, NOT to answer the question.\n\nRules:\n1. Include speaker attribution for each item. If unknown, write UNKNOWN.\n2. Prefer exact quotes, but if exact...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are extracting evidence from conversation excerpts for an adversarial question.\nYour job is to extract relevant evidence, NOT to answer the question.\n\nRules:\n1. Include speaker attribution for each item. If unknown, write UNKNOWN.\n2. Prefer exact quotes, but if exact...
[ { "role": "system", "content": "You are a memory compression expert. The question requires connecting multiple facts from different parts of the conversation.\n\nRules:\n1. Output a numbered list of factual observations.\n2. For each relevant fact, note its SOURCE (which excerpt or conversation segment it c...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: When was ML Pipeline last deployed?\n\nCandidates:\n[B-8429] Decision: Migrate to GitHub A...
[ { "role": "system", "content": "You are a memory compression expert. The question asks about timing, sequence, or chronological order of events.\n\nRules:\n1. Output a numbered list of factual observations in CHRONOLOGICAL ORDER.\n2. Include exact dates, times, and relative ordering (before/after/during).\n...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are a memory compression expert. The question asks about timing, sequence, or chronological order of events.\n\nRules:\n1. Output a numbered list of factual observations in CHRONOLOGICAL ORDER.\n2. Include exact dates, times, and relative ordering (before/after/during).\n...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: Who is responsible for Terraform maintenance?\n\nCandidates:\n[B-6463] Meeting notes from ...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are a memory compression expert. The question asks about timing, sequence, or chronological order of events.\n\nRules:\n1. Output a numbered list of factual observations in CHRONOLOGICAL ORDER.\n2. Include exact dates, times, and relative ordering (before/after/during).\n...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are a memory compression expert. The question asks about timing, sequence, or chronological order of events.\n\nRules:\n1. Output a numbered list of factual observations in CHRONOLOGICAL ORDER.\n2. Include exact dates, times, and relative ordering (before/after/during).\n...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are a memory compression expert. The question asks about timing, sequence, or chronological order of events.\n\nRules:\n1. Output a numbered list of factual observations in CHRONOLOGICAL ORDER.\n2. Include exact dates, times, and relative ordering (before/after/during).\n...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are a memory compression expert. Given a set of conversation memory excerpts and a question, extract ONLY the facts relevant to answering the question.\n\nRules:\n1. Output a numbered list of factual observations derived from the context.\n2. Each observation must be a si...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are a memory compression expert. The question asks about timing, sequence, or chronological order of events.\n\nRules:\n1. Output a numbered list of factual observations in CHRONOLOGICAL ORDER.\n2. Include exact dates, times, and relative ordering (before/after/during).\n...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: When was Mobile App v3 last deployed?\n\nCandidates:\n[B-3309] Ahmed discussed naestro-bot...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: How does mindlang.dev handle authentication?\n\nCandidates:\n[B-7895] Lisa from marketing ...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: When was mindlang.dev last deployed?\n\nCandidates:\n[B-4502] David Park reported a bug in...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: When was DataLake v2 last deployed?\n\nCandidates:\n[B-9511] Alice Chen reported a bug in ...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: When was CogNet last deployed?\n\nCandidates:\n[B-1891] Professor Garcia discussed ML Pipe...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: What did Alice Chen decide about React?\n\nCandidates:\n[B-6597] Meeting notes from 2026-0...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are extracting evidence from conversation excerpts for an adversarial question.\nYour job is to extract relevant evidence, NOT to answer the question.\n\nRules:\n1. Include speaker attribution for each item. If unknown, write UNKNOWN.\n2. Prefer exact quotes, but if exact...
[ { "role": "system", "content": "You are extracting evidence from conversation excerpts for an adversarial question.\nYour job is to extract relevant evidence, NOT to answer the question.\n\nRules:\n1. Include speaker attribution for each item. If unknown, write UNKNOWN.\n2. Prefer exact quotes, but if exact...
[ { "role": "system", "content": "You are a memory compression expert. The question requires connecting multiple facts from different parts of the conversation.\n\nRules:\n1. Output a numbered list of factual observations.\n2. For each relevant fact, note its SOURCE (which excerpt or conversation segment it c...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: When was naestro-bot last deployed?\n\nCandidates:\n[B-7395] InfraScale partnership update...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: Who is responsible for TensorFlow maintenance?\n\nCandidates:\n[B-3358] CTO Jamie Lee disc...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: What did Sarah decide about Supabase?\n\nCandidates:\n[B-4980] Meeting notes from 2026-02-...
[ { "role": "system", "content": "You are a memory compression expert. The question asks about timing, sequence, or chronological order of events.\n\nRules:\n1. Output a numbered list of factual observations in CHRONOLOGICAL ORDER.\n2. Include exact dates, times, and relative ordering (before/after/during).\n...
[ { "role": "system", "content": "You are extracting evidence from conversation excerpts for an adversarial question.\nYour job is to extract relevant evidence, NOT to answer the question.\n\nRules:\n1. Include speaker attribution for each item. If unknown, write UNKNOWN.\n2. Prefer exact quotes, but if exact...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are extracting evidence from conversation excerpts for an adversarial question.\nYour job is to extract relevant evidence, NOT to answer the question.\n\nRules:\n1. Include speaker attribution for each item. If unknown, write UNKNOWN.\n2. Prefer exact quotes, but if exact...
[ { "role": "system", "content": "You are a memory compression expert. Given a set of conversation memory excerpts and a question, extract ONLY the facts relevant to answering the question.\n\nRules:\n1. Output a numbered list of factual observations derived from the context.\n2. Each observation must be a si...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: Who is responsible for Django maintenance?\n\nCandidates:\n[B-3667] Sarah discussed DataLa...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are a memory compression expert. Given a set of conversation memory excerpts and a question, extract ONLY the facts relevant to answering the question.\n\nRules:\n1. Output a numbered list of factual observations derived from the context.\n2. Each observation must be a si...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: What did Professor Garcia decide about Kafka?\n\nCandidates:\n[B-9854] Ahmed discussed min...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are a memory compression expert. The question requires connecting multiple facts from different parts of the conversation.\n\nRules:\n1. Output a numbered list of factual observations.\n2. For each relevant fact, note its SOURCE (which excerpt or conversation segment it c...
[ { "role": "system", "content": "You are extracting evidence from conversation excerpts for an adversarial question.\nYour job is to extract relevant evidence, NOT to answer the question.\n\nRules:\n1. Include speaker attribution for each item. If unknown, write UNKNOWN.\n2. Prefer exact quotes, but if exact...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: How does API Gateway handle authentication?\n\nCandidates:\n[B-5016] Decision: Migrate to ...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: How does Edge Compute handle authentication?\n\nCandidates:\n[B-9109] Eve Johnson reported...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: How does DataLake v2 handle authentication?\n\nCandidates:\n[B-7317] Nikolai discussed Aut...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: How does 512-mind handle authentication?\n\nCandidates:\n[B-4097] Frank Williams reported ...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: How does API Gateway handle authentication?\n\nCandidates:\n[B-5618] Sarah reported a bug ...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: Who is responsible for FastAPI maintenance?\n\nCandidates:\n[B-7622] Decision: Migrate to ...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: How does NikolaChess handle authentication?\n\nCandidates:\n[B-3326] DataFlow Labs partner...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: When was naestro-bot last deployed?\n\nCandidates:\n[B-5589] Decision: Migrate to SQLAlche...
[ { "role": "system", "content": "You are Mind7B, a specialized memory reranking model. Given a query and candidate memory blocks, score each for relevance (0.0-1.0)." }, { "role": "user", "content": "Query: How does API Gateway handle authentication?\n\nCandidates:\n[B-5208] Meeting notes from 20...
[ { "role": "system", "content": "You are a memory compression expert. The question requires connecting multiple facts from different parts of the conversation.\n\nRules:\n1. Output a numbered list of factual observations.\n2. For each relevant fact, note its SOURCE (which excerpt or conversation segment it c...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract entities from text and return structured JSON." }, { "role": "user", "content": "Extract entities from the following text. Return a JSON array of objects, each with keys: \"name\" (string), \"type\...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...
[ { "role": "system", "content": "You are Mind7B, a specialized memory extraction model. You extract factual claims from text and return structured JSON." }, { "role": "user", "content": "Extract factual claims from the following text. Return a JSON array of objects, each with keys: \"claim\" (str...