{"schema_version":"onlylabs.public_signal.v1","title":"Together AI Model: togethercomputer/gemma-2-9b-it-MoAA-DPO","description":"Together AI model signal with public source context, captured evidence pages, related signals, and category-scoped analysis context.","url":"https://onlylabs.fyi/signals/f9ef9883-7f1e-493e-a1ac-a04be0da3d0d","json_url":"https://onlylabs.fyi/signals/f9ef9883-7f1e-493e-a1ac-a04be0da3d0d/signal.json","generated_at":"2026-06-11T03:04:00.642625+00:00","org":{"slug":"together-ai","name":"Together AI","category":"neocloud","category_label":"Neocloud","dossier_url":"https://onlylabs.fyi/labs/together-ai","dossier_json_url":"https://onlylabs.fyi/labs/together-ai/dossier.json"},"related_urls":{"signal":"https://onlylabs.fyi/signals/f9ef9883-7f1e-493e-a1ac-a04be0da3d0d","signal_json":"https://onlylabs.fyi/signals/f9ef9883-7f1e-493e-a1ac-a04be0da3d0d/signal.json","source":"https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO","lab_dossier":"https://onlylabs.fyi/labs/together-ai","lab_dossier_json":"https://onlylabs.fyi/labs/together-ai/dossier.json","analysis":"https://onlylabs.fyi/analysis/together-ai","analysis_json":"https://onlylabs.fyi/analysis/together-ai/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/together-ai/evidence.json","category":"https://onlylabs.fyi/neoclouds","category_json":"https://onlylabs.fyi/neoclouds.json","category_feed":"https://onlylabs.fyi/neoclouds/feed.xml","category_signals_json":"https://onlylabs.fyi/signals.json?category=neocloud","topic":"https://onlylabs.fyi/topics/releases","topic_signals_json":"https://onlylabs.fyi/topics/releases/signals.json?category=neocloud","topic_feed":"https://onlylabs.fyi/topics/releases/feed.xml?category=neocloud","data_business":null},"answer_pack":{"answer":"Together AI published togethercomputer/gemma-2-9b-it-MoAA-DPO. This model signal is evidence of what shipped on model infrastructure and how the release is positioned. High-signal details: 9 HF downloads · Low downloads, minor fine-tune. onlylabs links this event to 1 captured evidence page and 6 related model signals.","signal_desk":"releases","source_context":{"source_url":"https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO","source_host":"huggingface.co","occurred_at":"2024-12-06T01:28:39+00:00","first_seen_at":"2026-06-06T01:49:45.746189+00:00","date_source":"source","context":null},"context_markers":[{"label":"Lab","value":"Together AI","source":"signal"},{"label":"Signal desk","value":"releases","source":"signal"},{"label":"Source host","value":"huggingface.co","source":"source"},{"label":"Task","value":"text-generation","source":"model"},{"label":"Library","value":"transformers","source":"model"},{"label":"Downloads","value":"9","source":"traction"},{"label":"Likes","value":"0","source":"traction"},{"label":"Notability","value":"Low downloads, minor fine-tune","source":"signal"},{"label":"Watch term","value":"Eval methodology","source":"evidence"},{"label":"Watch term","value":"Model card","source":"model"},{"label":"Watch term","value":"Data pipeline","source":"evidence"},{"label":"Watch term","value":"Infrastructure","source":"evidence"},{"label":"Watch term","value":"Safety and alignment","source":"evidence"},{"label":"Watch term","value":"Agents and tool use","source":"evidence"}],"evidence_coverage":{"target_pages":1,"captured_pages":1,"readable_pages":1,"capture_methods":["plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO/raw/main/README.md"],"related_signals":6,"has_source_url":true,"latest_page_fetched_at":"2026-06-11T03:04:00.642625+00:00"},"data_business":{"matches":false,"lanes":[],"matched_terms":[],"score":null,"reason":null},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/f9ef9883-7f1e-493e-a1ac-a04be0da3d0d/signal.json","dossier_json":"https://onlylabs.fyi/labs/together-ai/dossier.json","analysis_json":"https://onlylabs.fyi/analysis/together-ai/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/together-ai/evidence.json","topic_signals_json":"https://onlylabs.fyi/topics/releases/signals.json?category=neocloud","topic_feed":"https://onlylabs.fyi/topics/releases/feed.xml?category=neocloud","category_signals_json":"https://onlylabs.fyi/signals.json?category=neocloud","data_radar_json":null,"opportunities_json":null},"analysis_playbook":{"objective":"Turn model cards, repository releases, and launch artifacts into a current view of what shipped and what changed.","evidence_focus":["model card","release note","version or tag","pipeline tag","license","downloads, likes, stars, and linked papers"],"extraction_questions":["What actually shipped, and through which public artifact?","Which releases include inspectable model, repo, paper, or package evidence?","Which launches create new eval, data, infra, or deployment needs?","Which shipped signals are missing captured page context?"],"signal_questions":["What shipped in this model artifact, and what evidence describes its capability, license, or traction?","What actually shipped, and through which public artifact?","Which releases include inspectable model, repo, paper, or package evidence?","Do the 6 related model signals show a repeated pattern?"],"output_fields":["org","release_type","artifact","traction","evidence_url"],"data_business_relevance":"Data-business lane extraction is scoped to frontier labs; for this category, keep conclusions tied to category-specific strategy, source evidence, and follow-up questions.","required_sources":[{"label":"signal_json","url":"https://onlylabs.fyi/signals/f9ef9883-7f1e-493e-a1ac-a04be0da3d0d/signal.json","required":true},{"label":"source","url":"https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO","required":true},{"label":"dossier_json","url":"https://onlylabs.fyi/labs/together-ai/dossier.json","required":true},{"label":"analysis_evidence_json","url":"https://onlylabs.fyi/analysis/together-ai/evidence.json","required":true},{"label":"topic_signals_json","url":"https://onlylabs.fyi/topics/releases/signals.json?category=neocloud","required":false},{"label":"data_radar_json","url":null,"required":false}],"expected_output":["one-paragraph source-grounded interpretation","category-specific implication","confidence and missing evidence","recommended next source to inspect"],"prompt_seed":"Using only the linked onlylabs JSON, captured source context, and cited evidence, analyze Together AI's model signal \"togethercomputer/gemma-2-9b-it-MoAA-DPO\" for neocloud strategy."},"semantic_triples":[{"subject":"Together AI","predicate":"published model","object":"togethercomputer/gemma-2-9b-it-MoAA-DPO","text":"Together AI published model togethercomputer/gemma-2-9b-it-MoAA-DPO."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"is classified as","object":"model signal","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO is classified as model signal."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"belongs to","object":"releases desk","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO belongs to releases desk."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has evidence coverage","object":"1 captured evidence page","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has evidence coverage 1 captured evidence page."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has captured page count","object":"1","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has captured page count 1."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has readable page count","object":"1","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has readable page count 1."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has related signal count","object":"6","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has related signal count 6."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has analysis playbook objective","object":"Turn model cards, repository releases, and launch artifacts into a current view of what shipped and what changed.","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has analysis playbook objective Turn model cards, repository releases, and launch artifacts into a current view of what shipped and what changed.."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has source host","object":"huggingface.co","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has source host huggingface.co."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has lab","object":"Together AI","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has lab Together AI."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has signal desk","object":"releases","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has signal desk releases."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has source host","object":"huggingface.co","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has source host huggingface.co."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has task","object":"text-generation","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has task text-generation."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has library","object":"transformers","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has library transformers."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has downloads","object":"9","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has downloads 9."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has likes","object":"0","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has likes 0."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has notability","object":"Low downloads, minor fine-tune","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has notability Low downloads, minor fine-tune."}]},"intelligence":{"signal_desk":"releases","answer":"Together AI published togethercomputer/gemma-2-9b-it-MoAA-DPO. This model signal is evidence of what shipped on model infrastructure and how the release is positioned. High-signal details: 9 HF downloads · Low downloads, minor fine-tune. onlylabs links this event to 1 captured evidence page and 6 related model signals.","semantic_triples":[{"subject":"Together AI","predicate":"published model","object":"togethercomputer/gemma-2-9b-it-MoAA-DPO","text":"Together AI published model togethercomputer/gemma-2-9b-it-MoAA-DPO."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"is classified as","object":"model signal","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO is classified as model signal."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"belongs to","object":"releases desk","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO belongs to releases desk."},{"subject":"togethercomputer/gemma-2-9b-it-MoAA-DPO","predicate":"has evidence coverage","object":"1 captured evidence page","text":"togethercomputer/gemma-2-9b-it-MoAA-DPO has evidence coverage 1 captured evidence page."}]},"signal":{"id":"f9ef9883-7f1e-493e-a1ac-a04be0da3d0d","url":"https://onlylabs.fyi/signals/f9ef9883-7f1e-493e-a1ac-a04be0da3d0d","json_url":"https://onlylabs.fyi/signals/f9ef9883-7f1e-493e-a1ac-a04be0da3d0d/signal.json","source_url":"https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO","title":"togethercomputer/gemma-2-9b-it-MoAA-DPO","summary":"Together AI published a model signal. onlylabs watches model cards, launch context, and traction.","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2024-12-06T01:28:39+00:00","first_seen_at":"2026-06-06T01:49:45.746189+00:00","date_source":"source","evidence_coverage":{"target_pages":1,"captured_pages":1,"readable_pages":1,"capture_methods":["plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO/raw/main/README.md"]},"facets":{"pipeline_tag":"text-generation"},"traction":{"github_stars":null,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":9,"hf_likes":0},"data_radar":null},"primary_evidence_page":{"url":"https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO/raw/main/README.md","final_url":"https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO/raw/main/README.md","title":"togethercomputer/gemma-2-9b-it-MoAA-DPO model card","http_status":200,"content_type":"text/plain; charset=utf-8","capture_method":"plain","fetched_at":"2026-06-11T03:04:00.642625+00:00","bytes":2690,"raw_path":"dd89477e7db8404a311da8b8edce30e8054356ebfa60c39ff1a4b389cc7c7bf7.md","content_hash":"447895621ff0663ce6f67955b224397c20e731e107d55048e106762bbe3c14bb","excerpt_chars":1200,"truncated":true,"excerpt":"--- library_name: transformers tags: [] --- Model Description This is the DPO model in our Mixture of Agents Alignment (MoAA) pipeline. This model is tuned on the Gemma-2-9b-it. MoAA is an approach that leverages collective intelligence from open‑source LLMs to advance alignment. Two mains stages are involved in our MoAA method. In the first stage, we employ MoA to produce high-quality synthetic data for supervised fine-tuning. In the second stage, we combines multiple LLMs as a reward model to provide preference annotations. Some key takeaways of our work: - 📈**Alignment pipeline that actually works** Our MoAA method sends Llama‑3.1‑8B‑Instruct’s Arena‑Hard **19 → 48** and Gemma-2-9B-it **42→56**, handily beating GPT‑4o‑labeled sets at the time. - 🏆**Ensembled rewards > single critics** An MoA reward model with dynamic criteria filtering edges out competitive ArmoRM on MT‑Bench & Arena‑Hard—all while staying 100 % open source. - 🚀**Self‑improvement unlocked** Fine‑tune the strongest model inside the ensemble on MoAA data and it *surpasses its own teachers*—evidence that open models can push past proprietary ceilings without external supervision. Model Sources For more details..."},"evidence_pages":[{"url":"https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO/raw/main/README.md","final_url":"https://huggingface.co/togethercomputer/gemma-2-9b-it-MoAA-DPO/raw/main/README.md","title":"togethercomputer/gemma-2-9b-it-MoAA-DPO model card","http_status":200,"content_type":"text/plain; charset=utf-8","capture_method":"plain","fetched_at":"2026-06-11T03:04:00.642625+00:00","bytes":2690,"raw_path":"dd89477e7db8404a311da8b8edce30e8054356ebfa60c39ff1a4b389cc7c7bf7.md","content_hash":"447895621ff0663ce6f67955b224397c20e731e107d55048e106762bbe3c14bb","excerpt_chars":1200,"truncated":true,"excerpt":"--- library_name: transformers tags: [] --- Model Description This is the DPO model in our Mixture of Agents Alignment (MoAA) pipeline. This model is tuned on the Gemma-2-9b-it. MoAA is an approach that leverages collective intelligence from open‑source LLMs to advance alignment. Two mains stages are involved in our MoAA method. In the first stage, we employ MoA to produce high-quality synthetic data for supervised fine-tuning. In the second stage, we combines multiple LLMs as a reward model to provide preference annotations. Some key takeaways of our work: - 📈**Alignment pipeline that actually works** Our MoAA method sends Llama‑3.1‑8B‑Instruct’s Arena‑Hard **19 → 48** and Gemma-2-9B-it **42→56**, handily beating GPT‑4o‑labeled sets at the time. - 🏆**Ensembled rewards > single critics** An MoA reward model with dynamic criteria filtering edges out competitive ArmoRM on MT‑Bench & Arena‑Hard—all while staying 100 % open source. - 🚀**Self‑improvement unlocked** Fine‑tune the strongest model inside the ensemble on MoAA data and it *surpasses its own teachers*—evidence that open models can push past proprietary ceilings without external supervision. Model Sources For more details..."}],"related_signals":[{"id":"22026f92-21da-461f-90e2-0debbedc287f","url":"https://onlylabs.fyi/signals/22026f92-21da-461f-90e2-0debbedc287f","source_url":"https://huggingface.co/togethercomputer/Aurora-Spec-Minimax-M2.5","title":"togethercomputer/Aurora-Spec-Minimax-M2.5","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-02-19T20:32:12+00:00","first_seen_at":"2026-06-06T01:49:45.746189+00:00","date_source":"source"},{"id":"0168c6b5-67ca-48c2-a176-2423a1f3f782","url":"https://onlylabs.fyi/signals/0168c6b5-67ca-48c2-a176-2423a1f3f782","source_url":"https://huggingface.co/togethercomputer/Aurora-Spec-Minimax-M2.1","title":"togethercomputer/Aurora-Spec-Minimax-M2.1","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-02-04T23:02:09+00:00","first_seen_at":"2026-06-06T01:49:45.746189+00:00","date_source":"source"},{"id":"57239096-e82d-4c02-b406-1d6717784602","url":"https://onlylabs.fyi/signals/57239096-e82d-4c02-b406-1d6717784602","source_url":"https://huggingface.co/togethercomputer/Aurora-Spec-Qwen3-Coder-Next-FP8","title":"togethercomputer/Aurora-Spec-Qwen3-Coder-Next-FP8","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-02-03T05:16:50+00:00","first_seen_at":"2026-06-06T01:49:45.746189+00:00","date_source":"source"},{"id":"93bbdc8d-2ee5-4111-99c5-34012f00fc79","url":"https://onlylabs.fyi/signals/93bbdc8d-2ee5-4111-99c5-34012f00fc79","source_url":"https://huggingface.co/togethercomputer/M1-3B","title":"togethercomputer/M1-3B","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2025-05-02T04:30:50+00:00","first_seen_at":"2026-06-06T01:49:45.746189+00:00","date_source":"source"},{"id":"1df1c794-a659-44cb-9190-bf7ccd672f04","url":"https://onlylabs.fyi/signals/1df1c794-a659-44cb-9190-bf7ccd672f04","source_url":"https://huggingface.co/togethercomputer/qwen2_vl_7b_instruct-biomedical-stage2","title":"togethercomputer/qwen2_vl_7b_instruct-biomedical-stage2","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2025-01-21T18:16:55+00:00","first_seen_at":"2026-06-06T01:49:45.746189+00:00","date_source":"source"},{"id":"2c3d3113-6f68-4140-8c03-6c54a29e967d","url":"https://onlylabs.fyi/signals/2c3d3113-6f68-4140-8c03-6c54a29e967d","source_url":"https://huggingface.co/togethercomputer/Llama-3.1-8B-Instruct-MoAA-DPO","title":"togethercomputer/Llama-3.1-8B-Instruct-MoAA-DPO","context":null,"kind":{"key":"model_released","label":"Model"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2024-12-05T23:47:39+00:00","first_seen_at":"2026-06-06T01:49:45.746189+00:00","date_source":"source"}]}