[{"lens":"arxiv","axis_key":"interestingness","axis_prompt":"Interestingness to a technically sophisticated reader tracking frontier AI research: how much the paper's core claim, if true, changes what researchers believe or build — weighing novelty of the idea, strength of evidence, and breadth of consequence over topical fashion or incremental gains.","axis_prompt_hash":"55076be568bc7da3a2435e16342d487ee435f8fbc3b43c97cd426183a3c1a328","prompt_variants":1,"entity_count":48,"last_scored_at":"2026-08-10T09:24:33.072000Z"},{"lens":"arxiv","axis_key":"practical-applicability","axis_prompt":"practical real-world applicability of the paper's contribution, as evidenced by the abstract","axis_prompt_hash":"ea1cc2a1265d21eb073b5ca689568be1c6c81a362927f6187f30d2c6b7929432","prompt_variants":1,"entity_count":15,"last_scored_at":"2026-08-07T04:17:47.185000Z"},{"lens":"hermes-agent-skills","axis_key":"scry-inspiration#a","axis_prompt":"Scry is a public, provenanced, SQL-queryable index of high-value internet corpora (papers, forums, social platforms, scholarly graphs) with metered query serving, vector search, and structured judgement campaigns, built and operated largely by autonomous agents. Which of these two agent skills is higher-leverage as INSPIRATION for Scry's own development — a capability pattern, workflow shape, or product idea that Scry's builders should study and adapt? Judge transferable design insight for a research-data-infrastructure product, not direct installability.","axis_prompt_hash":"1bf4892920e374f03cf0e58099267240c83a36468e658e695cf37f9af6eda183","prompt_variants":1,"entity_count":181,"last_scored_at":"2026-07-26T08:41:16.343000Z"},{"lens":"hermes-agent-skills","axis_key":"scry-inspiration#b","axis_prompt":"You maintain a corpus-indexing and research-query platform (bounded public SQL over crawled corpora, provenance-first schema design, agent-facing research workflows, usage-metered serving). Reading these two skill documents as design precedents, which one would teach your team more — richer patterns worth borrowing for the platform's crawling, indexing, query products, agent interfaces, or operations?","axis_prompt_hash":"cd7f077fc6ae49a5399f05c41fdedb8ee01fee27d2ced2762a87476dd3cfd3c3","prompt_variants":1,"entity_count":181,"last_scored_at":"2026-07-26T08:39:52.541000Z"},{"lens":"hermes-agent-skills","axis_key":"xphil-query-substrate#a","axis_prompt":"Which of these two agent skills is more valuable as part of a query substrate for existential philanthropy — capability infrastructure an analyst or research agent could invoke and compose when directing money, talent, and attention toward reducing existential risk and improving humanity's long-term trajectory? Weigh how much the skill helps locate, structure, verify, or act on decision-relevant information for such priorities; generic conveniences rank below skills that compound into serious research and grantmaking power.","axis_prompt_hash":"0337bf79ed3889ccee41d8115dd7d18a59d4a32fe8a0b7f990a3d63fa314b7c4","prompt_variants":1,"entity_count":181,"last_scored_at":"2026-07-26T08:36:54.513000Z"},{"lens":"hermes-agent-skills","axis_key":"xphil-query-substrate#b","axis_prompt":"An existential philanthropist is assembling a toolbox of agent skills to power the research and grantmaking agents behind their giving: mapping risks from advanced AI, biotechnology, and great-power conflict; finding and evaluating grantees; monitoring the information environment; and turning findings into funded action. Which of these two skills, installed in that toolbox, contributes more real capability to that mission?","axis_prompt_hash":"a957a64820855af3070b1742454571d333d5c0a95051a96593c0c2f310a19ddf","prompt_variants":1,"entity_count":181,"last_scored_at":"2026-07-26T08:38:18.359000Z"},{"lens":"landmark-ml-abstracts","axis_key":"empirical-claim-density","axis_prompt":"the density of concrete, falsifiable empirical claims in the text: specific measured results, named benchmarks, and quantitative comparisons a reader could verify, relative to the amount of text","axis_prompt_hash":"6f55d5d419e15f88a1a3493b7553ebe95677ec0b7eedeed90ffb384cf681d0ba","prompt_variants":1,"entity_count":24,"last_scored_at":"2026-08-07T03:43:19.164000Z"},{"lens":"landmark-ml-abstracts","axis_key":"self-containedness","axis_prompt":"the self-containedness and understandability of a piece of text: how fully it can be understood on its own, without outside context the reader does not have","axis_prompt_hash":"199209512d2e63a44fed2eae857bfff4a7c44446aeac81e34003350700258f03","prompt_variants":1,"entity_count":24,"last_scored_at":"2026-08-07T03:45:20.059000Z"},{"lens":"lesswrong","axis_key":"enduring-epistemic-value","axis_prompt":"Which of these two LessWrong posts has more enduring epistemic value — insight a careful reader would still profit from a decade after publication? Weigh durable conceptual contribution over topicality, style, or community significance. Higher means more enduring value.","axis_prompt_hash":"ba2b127c09acca0b44ab6f19311cec5db2829bb7b5fc6e89615e5c6fed540d88","prompt_variants":1,"entity_count":48,"last_scored_at":"2026-08-04T06:45:45.659000Z"},{"lens":"manifund","axis_key":"epistemic-integrity","axis_prompt":"epistemic integrity of the write-up: honest failure modes, quantified claims, falsifiable milestones","axis_prompt_hash":"1e88f6ebbbd8b052de402122aec2775107a748112fa69f9ee623408c8bf7a77d","prompt_variants":1,"entity_count":40,"last_scored_at":"2026-08-22T22:19:16.084000Z"},{"lens":"manifund","axis_key":"impact-per-marginal-dollar","axis_prompt":"expected impact per marginal dollar at the stated ask","axis_prompt_hash":"0766f47e15b1b3da8ac900ae915147ed16404bd797a0c39cf25a40073b25aac5","prompt_variants":1,"entity_count":40,"last_scored_at":"2026-08-22T22:19:40.232000Z"},{"lens":"manifund","axis_key":"team-track-record-evidence","axis_prompt":"verifiable track-record evidence the team can execute","axis_prompt_hash":"420dab4ba4e6663f850f9cccec31096b4721c333f4e04ac9c0207a5c6cd51975","prompt_variants":1,"entity_count":40,"last_scored_at":"2026-08-22T22:19:13.883000Z"},{"lens":"manifund","axis_key":"theory-of-change-plausibility","axis_prompt":"plausibility of the causal path from activities to claimed impact","axis_prompt_hash":"97744464398bbcbdafc1f0b759c246d17622ec659cbe70e218df9351edc785aa","prompt_variants":1,"entity_count":40,"last_scored_at":"2026-08-22T22:19:44.316000Z"},{"lens":"manifund-goals","axis_key":"importance-for-ai-safety-technical","axis_prompt":"importance of this evaluation criterion to a grantmaker whose aim is advancing technical AI safety research","axis_prompt_hash":"080477de42e2fa337127bcd64bfac0b377d2b5f2666170d39f503ff293c164e2","prompt_variants":1,"entity_count":4,"last_scored_at":"2026-08-20T03:38:20.539000Z"},{"lens":"manifund-goals","axis_key":"importance-for-careful-generalist","axis_prompt":"importance of this evaluation criterion to a grantmaker whose aim is a careful generalist grantmaker with no single ideology, who wants to fund what is actually good","axis_prompt_hash":"0690386652411fddbd4a00033132657d9c351264eb4093514ed50329806d95ac","prompt_variants":1,"entity_count":4,"last_scored_at":"2026-08-20T03:38:24.227000Z"},{"lens":"manifund-goals","axis_key":"importance-for-epistemic-public-good","axis_prompt":"importance of this evaluation criterion to a grantmaker whose aim is producing epistemic public goods: research, data, and tools that improve collective understanding","axis_prompt_hash":"77856e578b6003e334b439a776c42e6bd7eed011e32126a887a7d08c065ee3c0","prompt_variants":1,"entity_count":4,"last_scored_at":"2026-08-20T03:37:58.184000Z"},{"lens":"manifund-goals","axis_key":"importance-for-field-building","axis_prompt":"importance of this evaluation criterion to a grantmaker whose aim is building a healthy research field: talent pipelines, institutions, and community infrastructure","axis_prompt_hash":"2993e09eef14907304d70460fc85c1a445370acee35d822fe613382fe81aed66","prompt_variants":1,"entity_count":4,"last_scored_at":"2026-08-20T03:38:21.422000Z"},{"lens":"manifund-goals","axis_key":"importance-for-hits-based-leverage","axis_prompt":"importance of this evaluation criterion to a grantmaker whose aim is hits-based giving: accepting many failures to find rare outsized wins","axis_prompt_hash":"140ba00b7e799d948888321f18b700024a4df9da3a4e0f2288b1e1aa3b9eaa85","prompt_variants":1,"entity_count":4,"last_scored_at":"2026-08-20T03:38:01.178000Z"},{"lens":"manifund-goals","axis_key":"importance-for-nearterm-welfare","axis_prompt":"importance of this evaluation criterion to a grantmaker whose aim is maximizing near-term, measurable improvements in welfare","axis_prompt_hash":"caee2fc5cc0f977cedc8e561050d86b13a3755ec037c21a8067763d7a7732ecb","prompt_variants":1,"entity_count":4,"last_scored_at":"2026-08-20T03:37:56.230000Z"},{"lens":"manifund-goals","axis_key":"importance-for-xrisk-longtermist","axis_prompt":"importance of this evaluation criterion to a grantmaker whose aim is reducing existential risk to humanitys's long-term future","axis_prompt_hash":"88d08d161d9a5f67d15f44dc89598993a978c479f7e0e25d2cd8bec449c83380","prompt_variants":1,"entity_count":4,"last_scored_at":"2026-08-20T03:37:56.196000Z"},{"lens":"openpriors-selfcheck","axis_key":"self-containedness","axis_prompt":"the self-containedness and understandability of a piece of text: how fully it can be understood on its own, without outside context the reader does not have","axis_prompt_hash":"199209512d2e63a44fed2eae857bfff4a7c44446aeac81e34003350700258f03","prompt_variants":1,"entity_count":4,"last_scored_at":"2026-08-10T15:16:38.431000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"easy-speedup-gpt5-era#a","axis_prompt":"how strongly this item evidences a large, cheaply-obtained software performance win delivered by a GPT-5-era model","axis_prompt_hash":"60681162ca44f1a6c63c5091d3b5e5a389a243e0ae8b6c070453c424a4faa9f3","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:07:30.086000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"easy-speedup-gpt5-era#b","axis_prompt":"extent to which this shows a modern frontier model producing a dramatic software speedup with little human effort","axis_prompt_hash":"6388847b3c9031ae265e8504b50facdc9eab87dc9e69054e6f0179050f80eeff","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:07:54.080000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"goal-obsession-fidelity#a","axis_prompt":"how vividly this item shows an agent holding itself to a quantitative goal a human would have rounded off or abandoned","axis_prompt_hash":"3fa816729350d51099e60de86fdbbb6119bee594d443591cc05728d6c3928dea","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:17:58.619000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"goal-obsession-fidelity#b","axis_prompt":"extent to which this captures machine perfectionism: refusing to declare success while a measurable target remains unmet","axis_prompt_hash":"3b1da2c26e5dfcad8a50352774d9906ee2638e5b920137067e10ea63b4b292fb","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:18:23.192000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"legacy-rewrite-economics#a","axis_prompt":"how strongly this item suggests whole classes of legacy software libraries are now economically worth rewriting with AI agents","axis_prompt_hash":"2edd1a7db50c288f1f54f487ccae5325dacf41b9e85d2de12b05dda188ef0c56","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:20:08.439000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"legacy-rewrite-economics#b","axis_prompt":"extent to which this shifts the build-vs-adopt calculus: bespoke agent-written replacements beating established libraries","axis_prompt_hash":"0a58db7cd049c95c375860b92e245ac996259f2d3694d2560fba0784f3622863","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:20:47.521000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"long-horizon-endurance#a","axis_prompt":"how much this item demonstrates an AI agent sustaining productive autonomous work over many hours without human steering","axis_prompt_hash":"a8d61cac8e2489015d76e4c187ca8702a01d3afbf8872a1ec13497ed5d314ffe","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:16:55.658000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"long-horizon-endurance#b","axis_prompt":"degree to which this shows long-horizon agent endurance: unattended, self-directed effort measured in hours not minutes","axis_prompt_hash":"1aad961c199aa491060a6d23451c49b0ef4b36b99a38930bfb25023a78879c93","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:17:18.740000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"remarkable-agent-opportunity#a","axis_prompt":"how strongly this item evidences a remarkable, previously-impractical opportunity unlocked by autonomous AI agents","axis_prompt_hash":"90b0b302c863c8574f78e63505c79b61d791e21a3bdf3a1b5eb8ab3a99ade54d","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:06:21.030000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"remarkable-agent-opportunity#b","axis_prompt":"degree to which this demonstrates AI agents opening up striking new possibilities that were not practical before","axis_prompt_hash":"0d4e1ae8a574a598641b07aed66206ea7cde1e02104a24d0a525fb59dd0a8ce7","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:06:47.704000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"retellability#a","axis_prompt":"how likely a sharp engineer is to retell this item's story unprompted within a week","axis_prompt_hash":"27fd18aa6bde5dcd223cc431a245e018cff6474fc4a0f680953e2dd820dbc8e3","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:21:15.291000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"retellability#b","axis_prompt":"degree to which this item is a memorable, repeatable anecdote rather than forgettable feed content","axis_prompt_hash":"9e1a6176a6210cc3154707ec452650522b22256b3434e55697027353ab49d28a","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:21:37.778000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"verifiable-performance-claim#a","axis_prompt":"how checkable the headline claim in this item is: benchmark, artifact, or reproducible harness attached rather than vibes","axis_prompt_hash":"992d2a689be7902ce5798a459f80ae58da9e94a5ae7ac5de4d4b802cf532e37a","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:18:48.189000Z"},{"lens":"remarkable-agent-outcomes","axis_key":"verifiable-performance-claim#b","axis_prompt":"degree to which a skeptical engineer could verify this item's central claim from evidence it points to","axis_prompt_hash":"73a65635147dcdb649203cd51f029a193abadec58549fcd3b013a1d4ba24ff80","prompt_variants":1,"entity_count":21,"last_scored_at":"2026-07-20T05:19:38.222000Z"}]