{"schema_version":"onlylabs.public_signal.v1","title":"AI21 Labs Writing: Improving Best Of N With Budget Aware Execution For Swe Agents","description":"AI21 Labs writing signal with public source context, captured evidence pages, related signals, and category-scoped analysis context.","url":"https://onlylabs.fyi/signals/f8d81bf1-81cf-4b75-8848-0f5176e42123","json_url":"https://onlylabs.fyi/signals/f8d81bf1-81cf-4b75-8848-0f5176e42123/signal.json","generated_at":"2026-07-26T20:41:43.570Z","evidence_latest_fetched_at":"2026-07-08T12:02:55.21834+00:00","signal_first_seen_at":"2026-07-08T12:01:45.257496+00:00","org":{"slug":"ai21","name":"AI21 Labs","category":"neolab","category_label":"Neolab","dossier_url":"https://onlylabs.fyi/labs/ai21","dossier_json_url":"https://onlylabs.fyi/labs/ai21/dossier.json"},"related_urls":{"signal":"https://onlylabs.fyi/signals/f8d81bf1-81cf-4b75-8848-0f5176e42123","signal_json":"https://onlylabs.fyi/signals/f8d81bf1-81cf-4b75-8848-0f5176e42123/signal.json","source":"https://www.ai21.com/blog/improving-best-of-n-with-budget-aware-execution-for-swe-agents/","lab_dossier":"https://onlylabs.fyi/labs/ai21","lab_dossier_json":"https://onlylabs.fyi/labs/ai21/dossier.json","analysis":"https://onlylabs.fyi/analysis/ai21","analysis_json":"https://onlylabs.fyi/analysis/ai21/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/ai21/evidence.json","category":"https://onlylabs.fyi/neolabs","category_json":"https://onlylabs.fyi/neolabs.json","category_feed":"https://onlylabs.fyi/neolabs/feed.xml","category_signals_json":"https://onlylabs.fyi/signals.json?category=neolab","topic":"https://onlylabs.fyi/topics/talking","topic_signals_json":"https://onlylabs.fyi/topics/talking/signals.json?category=neolab","topic_feed":"https://onlylabs.fyi/topics/talking/feed.xml?category=neolab","data_business":null},"answer_pack":{"answer":"AI21 Labs published Improving Best Of N With Budget Aware Execution For Swe Agents. This talking signal gives public context for research themes, product direction, policy, or launch framing. High-signal details: Improving Best-of-N with Budget-Aware Execution for SWE Agents | AI21 Skip to Main Menu Skip to Main Content Skip to Footer Back to Blog --> Back to Blog In brief As.... onlylabs links this event to 1 captured evidence page and 6 related writing signals.","signal_desk":"talking","source_context":{"source_url":"https://www.ai21.com/blog/improving-best-of-n-with-budget-aware-execution-for-swe-agents/","source_host":"ai21.com","occurred_at":"2026-07-08T10:48:05+00:00","first_seen_at":"2026-07-08T12:01:45.257496+00:00","date_source":"sitemap.lastmod","context":null},"context_markers":[{"label":"Lab","value":"AI21 Labs","source":"signal"},{"label":"Signal desk","value":"talking","source":"signal"},{"label":"Source host","value":"ai21.com","source":"source"},{"label":"Watch term","value":"Eval methodology","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://www.ai21.com/blog/improving-best-of-n-with-budget-aware-execution-for-swe-agents/"],"related_signals":6,"has_source_url":true,"latest_page_fetched_at":"2026-07-08T12:02:55.21834+00:00"},"data_business":{"matches":false,"lanes":[],"matched_terms":[],"score":null,"reason":null},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/f8d81bf1-81cf-4b75-8848-0f5176e42123/signal.json","dossier_json":"https://onlylabs.fyi/labs/ai21/dossier.json","analysis_json":"https://onlylabs.fyi/analysis/ai21/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/ai21/evidence.json","topic_signals_json":"https://onlylabs.fyi/topics/talking/signals.json?category=neolab","topic_feed":"https://onlylabs.fyi/topics/talking/feed.xml?category=neolab","category_signals_json":"https://onlylabs.fyi/signals.json?category=neolab","data_radar_json":null,"opportunities_json":null},"analysis_playbook":{"objective":"Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.","evidence_focus":["post title","source URL","captured page text","HN traction","linked model or paper references","publication date"],"extraction_questions":["Which themes are labs choosing to explain publicly?","Which posts are attracting outside discussion?","Which writing reframes a recent release, model, hiring wave, or policy stance?","Which posts mention data, evals, infrastructure, safety, or deployment workflows?"],"signal_questions":["What public theme, launch framing, or research direction does this writing signal expose?","Which themes are labs choosing to explain publicly?","Which posts are attracting outside discussion?","Do the 6 related writing signals show a repeated pattern?"],"output_fields":["org","theme","public_framing","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/f8d81bf1-81cf-4b75-8848-0f5176e42123/signal.json","required":true},{"label":"source","url":"https://www.ai21.com/blog/improving-best-of-n-with-budget-aware-execution-for-swe-agents/","required":true},{"label":"dossier_json","url":"https://onlylabs.fyi/labs/ai21/dossier.json","required":true},{"label":"analysis_evidence_json","url":"https://onlylabs.fyi/analysis/ai21/evidence.json","required":true},{"label":"topic_signals_json","url":"https://onlylabs.fyi/topics/talking/signals.json?category=neolab","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 AI21 Labs's writing signal \"Improving Best Of N With Budget Aware Execution For Swe Agents\" for neolab strategy."},"semantic_triples":[{"subject":"AI21 Labs","predicate":"published","object":"Improving Best Of N With Budget Aware Execution For Swe Agents","text":"AI21 Labs published Improving Best Of N With Budget Aware Execution For Swe Agents."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"is classified as","object":"writing signal","text":"Improving Best Of N With Budget Aware Execution For Swe Agents is classified as writing signal."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"belongs to","object":"talking desk","text":"Improving Best Of N With Budget Aware Execution For Swe Agents belongs to talking desk."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has evidence coverage","object":"1 captured evidence page","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has evidence coverage 1 captured evidence page."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has captured page count","object":"1","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has captured page count 1."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has readable page count","object":"1","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has readable page count 1."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has related signal count","object":"6","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has related signal count 6."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has analysis playbook objective","object":"Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has analysis playbook objective Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has source host","object":"ai21.com","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has source host ai21.com."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has lab","object":"AI21 Labs","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has lab AI21 Labs."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has signal desk","object":"talking","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has signal desk talking."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has source host","object":"ai21.com","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has source host ai21.com."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has watch term","object":"Eval methodology","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has watch term Eval methodology."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has watch term","object":"Infrastructure","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has watch term Infrastructure."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has watch term","object":"Safety and alignment","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has watch term Safety and alignment."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has watch term","object":"Agents and tool use","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has watch term Agents and tool use."}]},"intelligence":{"signal_desk":"talking","answer":"AI21 Labs published Improving Best Of N With Budget Aware Execution For Swe Agents. This talking signal gives public context for research themes, product direction, policy, or launch framing. High-signal details: Improving Best-of-N with Budget-Aware Execution for SWE Agents | AI21 Skip to Main Menu Skip to Main Content Skip to Footer Back to Blog --> Back to Blog In brief As.... onlylabs links this event to 1 captured evidence page and 6 related writing signals.","semantic_triples":[{"subject":"AI21 Labs","predicate":"published","object":"Improving Best Of N With Budget Aware Execution For Swe Agents","text":"AI21 Labs published Improving Best Of N With Budget Aware Execution For Swe Agents."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"is classified as","object":"writing signal","text":"Improving Best Of N With Budget Aware Execution For Swe Agents is classified as writing signal."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"belongs to","object":"talking desk","text":"Improving Best Of N With Budget Aware Execution For Swe Agents belongs to talking desk."},{"subject":"Improving Best Of N With Budget Aware Execution For Swe Agents","predicate":"has evidence coverage","object":"1 captured evidence page","text":"Improving Best Of N With Budget Aware Execution For Swe Agents has evidence coverage 1 captured evidence page."}]},"signal":{"id":"f8d81bf1-81cf-4b75-8848-0f5176e42123","url":"https://onlylabs.fyi/signals/f8d81bf1-81cf-4b75-8848-0f5176e42123","json_url":"https://onlylabs.fyi/signals/f8d81bf1-81cf-4b75-8848-0f5176e42123/signal.json","source_url":"https://www.ai21.com/blog/improving-best-of-n-with-budget-aware-execution-for-swe-agents/","title":"Improving Best Of N With Budget Aware Execution For Swe Agents","summary":"AI21 Labs published a writing signal. onlylabs watches public writing for research themes, product direction, and model-launch context.","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2026-07-08T10:48:05+00:00","first_seen_at":"2026-07-08T12:01:45.257496+00:00","date_source":"sitemap.lastmod","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://www.ai21.com/blog/improving-best-of-n-with-budget-aware-execution-for-swe-agents/"]},"facets":{},"traction":{"github_stars":null,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":null,"hf_likes":null},"data_radar":null},"primary_evidence_page":{"is_primary":true,"source_match":true,"url":"https://www.ai21.com/blog/improving-best-of-n-with-budget-aware-execution-for-swe-agents/","final_url":"https://www.ai21.com/blog/improving-best-of-n-with-budget-aware-execution-for-swe-agents/","title":"Improving Best Of N With Budget Aware Execution For Swe Agents","http_status":200,"content_type":"text/html; charset=UTF-8","capture_method":"plain","fetched_at":"2026-07-08T12:02:55.21834+00:00","bytes":76032,"raw_path":"da03097fb92fa7ee0ba184f2a1beee2e6cc0ee51c1a9c942b30407322d82d669.html","content_hash":"11976212136b8dfa7e07984fdc228dd0cbbe82c2e8bcc2ed70d9d3ec371b75cb","excerpt_chars":1200,"truncated":true,"excerpt":"Improving Best-of-N with Budget-Aware Execution for SWE Agents | AI21 Skip to Main Menu Skip to Main Content Skip to Footer Back to Blog --> Back to Blog In brief As part of our broader focus on agent optimization, we’ve released a study demonstrating the efficacy of both horizontal and vertical scaling strategies for SWE agents. However, in our experiments, we have traditionally applied a uniform compute budget across problems, as if every task shares the same difficulty. But outside of a controlled lab setting, that assumption doesn’t hold up. In reality, we know that task difficulty follows a heavy-tailed distribution: a meaningful fraction of tasks are solved on low effort, while others require a large budget. For example, in a previous post detailing our efforts reaching a state-of-the-art result on SWE-rebench , we found that roughly 50% of tasks sampled from that benchmark are resolved by a single rollout of GPT-5.2, yet our original policy was spending five rollouts on every single one of them. That’s money being left on the table. At a moment when so many companies are examining their token consumption and looking for ways to eliminate cost, we present our latest..."},"evidence_pages":[],"related_signals":[{"id":"5831d1b3-d2e5-432e-9dc5-7cd00f90893c","url":"https://onlylabs.fyi/signals/5831d1b3-d2e5-432e-9dc5-7cd00f90893c","source_url":"https://www.ai21.com/blog/better-and-cheaper-together-open-models-explore-frontier-models-patch/","title":"Better And Cheaper Together Open Models Explore Frontier Models Patch","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2026-07-16T09:45:10+00:00","first_seen_at":"2026-07-16T12:00:49.961598+00:00","date_source":"sitemap.lastmod"},{"id":"9046dede-d89f-4925-80a3-7ea01b0df483","url":"https://onlylabs.fyi/signals/9046dede-d89f-4925-80a3-7ea01b0df483","source_url":"https://www.ai21.com/blog/spend-isnt-going-down-what-now/","title":"Spend Isnt Going Down What Now","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2026-06-30T18:12:31+00:00","first_seen_at":"2026-06-26T20:26:30.018378+00:00","date_source":"sitemap.lastmod"},{"id":"e37793eb-b716-448c-861e-93b1bcc4d62b","url":"https://onlylabs.fyi/signals/e37793eb-b716-448c-861e-93b1bcc4d62b","source_url":"https://www.ai21.com/blog/merging-weak-agents-into-a-state-of-the-art-deep-researcher/","title":"Merging Weak Agents Into A State Of The Art Deep Researcher","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2026-06-24T00:00:00.000Z","first_seen_at":"2026-06-26T20:26:30.018378+00:00","date_source":"page.visible_date"},{"id":"8b7c7a55-a3c5-4302-bfe6-bd2991b6df06","url":"https://onlylabs.fyi/signals/8b7c7a55-a3c5-4302-bfe6-bd2991b6df06","source_url":"https://www.ai21.com/blog/first-scale-then-enrich-how-the-right-execution-strategy-helped-us-reach-state-of-the-art-on-swe-rebench/","title":"First Scale Then Enrich How The Right Execution Strategy Helped Us Reach State Of The Art On Swe Rebench","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2026-06-07T10:19:21+00:00","first_seen_at":"2026-06-26T20:26:30.018378+00:00","date_source":"sitemap.lastmod"},{"id":"52ae8840-7ad4-4231-977d-c4440f553d42","url":"https://onlylabs.fyi/signals/52ae8840-7ad4-4231-977d-c4440f553d42","source_url":"https://www.ai21.com/blog/caching-in-agentic-llm-pipelines/","title":"Caching In Agentic Llm Pipelines","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2026-05-14T12:00:57+00:00","first_seen_at":"2026-06-26T20:26:30.018378+00:00","date_source":"sitemap.lastmod"},{"id":"ac47d724-0b12-4434-9be6-d99a6fb3423f","url":"https://onlylabs.fyi/signals/ac47d724-0b12-4434-9be6-d99a6fb3423f","source_url":"https://www.ai21.com/blog/ai-system-requirements/","title":"Ai System Requirements","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"ai21","name":"AI21 Labs","category":"neolab"},"occurred_at":"2026-05-07T11:11:22+00:00","first_seen_at":"2026-06-26T20:26:30.018378+00:00","date_source":"sitemap.lastmod"}]}