[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-innovative-method-to-identify-scientific-breakthroughs-in-research-history-en":3,"ArticleBody_hZYBcRMyNfQGAvM0vIl7qg55EnJat5udbFhgWdFBGk":185},{"article":4,"relatedArticles":177,"locale":62},{"id":5,"title":6,"slug":7,"content":8,"htmlContent":9,"excerpt":10,"category":11,"tags":12,"metaDescription":10,"wordCount":13,"readingTime":14,"publishedAt":15,"sources":16,"sourceCoverage":54,"transparency":56,"seo":59,"language":62,"featuredImage":63,"featuredImageCredit":64,"isFreeGeneration":68,"trendSlug":69,"niche":70,"geoTakeaways":74,"geoFaq":83,"entities":93},"69d18d5423cc38232fa5c574","Innovative method to identify scientific breakthroughs in research history","innovative-method-to-identify-scientific-breakthroughs-in-research-history","## Introduction\n\nScience history is usually told through landmark discoveries like evolution, atomic fission, and antibiotics. [2][3]  \nBut until recently, there was no scalable way to scan the full research record and identify which papers actually redirected the course of science.\n\nA team led by Sadamori Kojaku at [Binghamton University](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FBinghamton_University), with collaborators at the [University of Virginia](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FUniversity_of_Virginia), created a method that maps about 55 million papers and patents to detect disruptive innovations. [1][3]  \nPublished in *[Science Advances](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FScience_Advances)* in April 2026, it provides a new way to track how breakthroughs emerge and spread. [2]\n\n💡 **Key takeaway:** Instead of counting citations, the method measures whether future work is pulled away from a paper’s predecessors—an operational definition of a “breakthrough.” [1][3]\n\nThis article outlines how the method works, how it connects to research on the birth of new fields, and what it implies for funding, strategy, and evaluation. [1][2][4]\n\n## Main Content\n\n### Key point 1: From counting citations to mapping disruption\n\nTraditional metrics like [citation counts](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FCitation) and impact factors: [3]\n\n- Measure how often a paper is cited  \n- Emphasize direct follow‑on work  \n- Capture visibility but often miss paradigm‑shifting research that makes prior work less central [3]\n\nThe new method uses [neural embedding](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FWord_embedding), representing each paper or patent as points in a high‑dimensional space. [1][3]  \nEach work receives **two** vectors:\n\n- A “past” vector summarizing the work it builds on  \n- A “future” vector summarizing the work that cites it [3]\n\nTheir difference captures disruptiveness:\n\n- Large divergence: future research clusters away from the paper’s own foundations (high disruption)  \n- Small divergence: future research stays aligned with prior work (incremental advance) [3]\n\nNobel‑level discoveries typically show especially large gaps between past and future vectors, consistent with launching new directions or fields. [3]\n\n📊 **Data point:** The team applied this dual‑vector model to ~55 million papers and patents, tracing disruptive events across modern research. [1][3]\n\nIn effect, the method distinguishes routine extensions from contributions that become new focal points for later research. [1][3]\n\n### Key point 2: Revealing hidden and simultaneous breakthroughs\n\nThe embedding approach can detect **simultaneous breakthroughs**—multiple groups independently converging on similar transformative ideas. [1][3]  \n\n- Traditional metrics scatter credit among these works, masking the collective shift. [3]  \n- Embeddings show when several papers jointly redirect citation flows into a new region of “idea space.” [1][3]\n\nThis is crucial in fast‑moving domains, such as [data‑intensive astronomy](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FList_of_software_for_astronomy_research_and_education), where facilities like the [Vera C. Rubin Observatory](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FVera_C._Rubin_Observatory) will generate more data in a year than all previous optical surveys combined. [5][9]\n\n💼 **Example:**  \nA small national agency might spot mid‑sized cancer immunotherapy labs whose papers share a disruptive “turn” in embedding space around specific techniques or biomarkers, even without standout citation counts. [1][3]\n\nThis view dovetails with studies of how new scientific fields arise. [4]\n\n- An analysis of 350+ fields found most are triggered by powerful methods or tools (e.g., advanced telescopes, x‑ray crystallography, [randomized trials](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FRandomized_controlled_trial)). [4]  \n- About a quarter of fields are essentially new methods themselves (e.g., [laser physics](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FLaser_science), [econometrics](https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FEconometrics)). [4]\n\n⚡ **Key point:** Methods that shift embedding trajectories and spawn new clusters are often the very tools that seed new fields. [1][3][4]\n\n### Key point 3: Implications for policy, evaluation, and practice\n\nFor science policy, disruptiveness offers a broader measure of impact: [1][2]\n\n- Focuses on whether work redirects future citations, not just how many it accumulates [1][3]  \n- Helps funders see if programs are opening new directions, even before citation counts peak\n\nA program officer could, for example:\n\n- Track whether high‑risk grants generate new embedding clusters  \n- Balance portfolios between steady, incremental output and high‑disruption bets [1][2]\n\nFor researchers, the method can provide: [1][3][4]\n\n- Clarity on how their work fits into long‑term trajectories  \n- Early signals of emerging methods becoming focal points  \n- Historical maps of past disruptive shifts in their field\n\n⚠️ **Key point:** Disruptiveness does not “prove” importance; it quantifies redirection patterns and must be interpreted with: [1][2][4]\n\n- Peer review and domain expertise  \n- Replication and robustness evidence\n\nLimitations include: [1][2][3][4]\n\n- Sensitivity of neural embeddings to training choices and database coverage  \n- Possible underestimation of specialized but socially crucial work  \n- Bias against research in underindexed languages and venues\n\n## Conclusion\n\n### Summary\n\nKojaku and colleagues’ method uses neural embeddings of both the intellectual past and future influence of each paper; their divergence becomes a disruptiveness score. [1][3]  \nApplied to tens of millions of papers and patents, it highlights iconic breakthroughs and simultaneous, distributed innovations that conventional citation metrics often overlook. [1][2][3]\n\nCombined with evidence that new fields usually emerge from powerful tools and methods, this approach quantitatively traces how such tools reshape research over time. [4]\n\n💡 **Key takeaway:** Breakthroughs appear not just as highly cited works, but as inflection points where the direction of science bends. [1][3][4]\n\n### Next steps (call to action)\n\nTo make use of this methodology:\n\n- **Funders** should pilot disruptiveness metrics in portfolio reviews and in programs aimed at transformative tools. [1][2][4]  \n- **Researchers** can mine disruption maps to spot underexplored methodological niches and learn from past field‑forming moments. [1][3][4]  \n- **Science‑of‑science scholars** should combine embeddings with qualitative case studies to understand when and why disruptive shifts succeed or stall. [1][2][4]\n\nThe broader goal is to use this method not as a ranking device, but as a navigational chart—guiding the scientific community in cultivating the methods and environments where tomorrow’s breakthroughs are most likely to emerge. [1][2][4]","\u003Ch2>Introduction\u003C\u002Fh2>\n\u003Cp>Science history is usually told through landmark discoveries like evolution, atomic fission, and antibiotics. \u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003Cbr>\nBut until recently, there was no scalable way to scan the full research record and identify which papers actually redirected the course of science.\u003C\u002Fp>\n\u003Cp>A team led by Sadamori Kojaku at \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FBinghamton_University\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">Binghamton University\u003C\u002Fa>, with collaborators at the \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FUniversity_of_Virginia\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">University of Virginia\u003C\u002Fa>, created a method that maps about 55 million papers and patents to detect disruptive innovations. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003Cbr>\nPublished in \u003Cem>\u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FScience_Advances\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">Science Advances\u003C\u002Fa>\u003C\u002Fem> in April 2026, it provides a new way to track how breakthroughs emerge and spread. \u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>💡 \u003Cstrong>Key takeaway:\u003C\u002Fstrong> Instead of counting citations, the method measures whether future work is pulled away from a paper’s predecessors—an operational definition of a “breakthrough.” \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>This article outlines how the method works, how it connects to research on the birth of new fields, and what it implies for funding, strategy, and evaluation. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fp>\n\u003Ch2>Main Content\u003C\u002Fh2>\n\u003Ch3>Key point 1: From counting citations to mapping disruption\u003C\u002Fh3>\n\u003Cp>Traditional metrics like \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FCitation\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">citation counts\u003C\u002Fa> and impact factors: \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Measure how often a paper is cited\u003C\u002Fli>\n\u003Cli>Emphasize direct follow‑on work\u003C\u002Fli>\n\u003Cli>Capture visibility but often miss paradigm‑shifting research that makes prior work less central \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>The new method uses \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FWord_embedding\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">neural embedding\u003C\u002Fa>, representing each paper or patent as points in a high‑dimensional space. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003Cbr>\nEach work receives \u003Cstrong>two\u003C\u002Fstrong> vectors:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>A “past” vector summarizing the work it builds on\u003C\u002Fli>\n\u003Cli>A “future” vector summarizing the work that cites it \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Their difference captures disruptiveness:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Large divergence: future research clusters away from the paper’s own foundations (high disruption)\u003C\u002Fli>\n\u003Cli>Small divergence: future research stays aligned with prior work (incremental advance) \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Nobel‑level discoveries typically show especially large gaps between past and future vectors, consistent with launching new directions or fields. \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>📊 \u003Cstrong>Data point:\u003C\u002Fstrong> The team applied this dual‑vector model to ~55 million papers and patents, tracing disruptive events across modern research. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>In effect, the method distinguishes routine extensions from contributions that become new focal points for later research. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Ch3>Key point 2: Revealing hidden and simultaneous breakthroughs\u003C\u002Fh3>\n\u003Cp>The embedding approach can detect \u003Cstrong>simultaneous breakthroughs\u003C\u002Fstrong>—multiple groups independently converging on similar transformative ideas. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Traditional metrics scatter credit among these works, masking the collective shift. \u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>Embeddings show when several papers jointly redirect citation flows into a new region of “idea space.” \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>This is crucial in fast‑moving domains, such as \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FList_of_software_for_astronomy_research_and_education\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">data‑intensive astronomy\u003C\u002Fa>, where facilities like the \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FVera_C._Rubin_Observatory\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">Vera C. Rubin Observatory\u003C\u002Fa> will generate more data in a year than all previous optical surveys combined. \u003Ca href=\"#source-5\" class=\"citation-link\" title=\"View source [5]\">[5]\u003C\u002Fa>\u003Ca href=\"#source-9\" class=\"citation-link\" title=\"View source [9]\">[9]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>💼 \u003Cstrong>Example:\u003C\u002Fstrong>\u003Cbr>\nA small national agency might spot mid‑sized cancer immunotherapy labs whose papers share a disruptive “turn” in embedding space around specific techniques or biomarkers, even without standout citation counts. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>This view dovetails with studies of how new scientific fields arise. \u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>An analysis of 350+ fields found most are triggered by powerful methods or tools (e.g., advanced telescopes, x‑ray crystallography, \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FRandomized_controlled_trial\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">randomized trials\u003C\u002Fa>). \u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>About a quarter of fields are essentially new methods themselves (e.g., \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FLaser_science\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">laser physics\u003C\u002Fa>, \u003Ca href=\"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FEconometrics\" class=\"wiki-link\" target=\"_blank\" rel=\"noopener\">econometrics\u003C\u002Fa>). \u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>⚡ \u003Cstrong>Key point:\u003C\u002Fstrong> Methods that shift embedding trajectories and spawn new clusters are often the very tools that seed new fields. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fp>\n\u003Ch3>Key point 3: Implications for policy, evaluation, and practice\u003C\u002Fh3>\n\u003Cp>For science policy, disruptiveness offers a broader measure of impact: \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Focuses on whether work redirects future citations, not just how many it accumulates \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>Helps funders see if programs are opening new directions, even before citation counts peak\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>A program officer could, for example:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Track whether high‑risk grants generate new embedding clusters\u003C\u002Fli>\n\u003Cli>Balance portfolios between steady, incremental output and high‑disruption bets \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>For researchers, the method can provide: \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Clarity on how their work fits into long‑term trajectories\u003C\u002Fli>\n\u003Cli>Early signals of emerging methods becoming focal points\u003C\u002Fli>\n\u003Cli>Historical maps of past disruptive shifts in their field\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>⚠️ \u003Cstrong>Key point:\u003C\u002Fstrong> Disruptiveness does not “prove” importance; it quantifies redirection patterns and must be interpreted with: \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Peer review and domain expertise\u003C\u002Fli>\n\u003Cli>Replication and robustness evidence\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Limitations include: \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Sensitivity of neural embeddings to training choices and database coverage\u003C\u002Fli>\n\u003Cli>Possible underestimation of specialized but socially crucial work\u003C\u002Fli>\n\u003Cli>Bias against research in underindexed languages and venues\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2>Conclusion\u003C\u002Fh2>\n\u003Ch3>Summary\u003C\u002Fh3>\n\u003Cp>Kojaku and colleagues’ method uses neural embeddings of both the intellectual past and future influence of each paper; their divergence becomes a disruptiveness score. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003Cbr>\nApplied to tens of millions of papers and patents, it highlights iconic breakthroughs and simultaneous, distributed innovations that conventional citation metrics often overlook. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>Combined with evidence that new fields usually emerge from powerful tools and methods, this approach quantitatively traces how such tools reshape research over time. \u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>💡 \u003Cstrong>Key takeaway:\u003C\u002Fstrong> Breakthroughs appear not just as highly cited works, but as inflection points where the direction of science bends. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fp>\n\u003Ch3>Next steps (call to action)\u003C\u002Fh3>\n\u003Cp>To make use of this methodology:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Funders\u003C\u002Fstrong> should pilot disruptiveness metrics in portfolio reviews and in programs aimed at transformative tools. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Researchers\u003C\u002Fstrong> can mine disruption maps to spot underexplored methodological niches and learn from past field‑forming moments. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-3\" class=\"citation-link\" title=\"View source [3]\">[3]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Science‑of‑science scholars\u003C\u002Fstrong> should combine embeddings with qualitative case studies to understand when and why disruptive shifts succeed or stall. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>The broader goal is to use this method not as a ranking device, but as a navigational chart—guiding the scientific community in cultivating the methods and environments where tomorrow’s breakthroughs are most likely to emerge. \u003Ca href=\"#source-1\" class=\"citation-link\" title=\"View source [1]\">[1]\u003C\u002Fa>\u003Ca href=\"#source-2\" class=\"citation-link\" title=\"View source [2]\">[2]\u003C\u002Fa>\u003Ca href=\"#source-4\" class=\"citation-link\" title=\"View source [4]\">[4]\u003C\u002Fa>\u003C\u002Fp>\n","Introduction\n\nScience history is usually told through landmark discoveries like evolution, atomic fission, and antibiotics. [2][3]  \nBut until recently, there was no scalable way to scan the full rese...","trend-radar",[],902,5,"2026-04-04T22:22:17.938Z",[17,22,26,30,34,38,42,46,50],{"title":18,"url":19,"summary":20,"type":21},"New Method Maps 55M Papers and Patents to Identify Disruptive Innovations","https:\u002F\u002Fwww.linkedin.com\u002Fposts\u002Fphys-org_a-new-way-to-detect-breakthroughs-in-science-activity-7445168662300344320-RGtN","A new analytical method has been developed to identify disruptive innovations in science and technology by mapping the influence of approximately 55 million papers and patents. Using neural embedding,...","kb",{"title":23,"url":24,"summary":25,"type":21},"New method pinpoints scientific breakthroughs, aiding research and funding - Journior","https:\u002F\u002Fnews.journior.com\u002Fagencynews\u002Fa-new-way-to-detect-breakthroughs-in-science-large-scale-analysis-reveals-disruptive-innovations-in-research-history-l4n93fiy\u002F","A new way to detect breakthroughs in science: Large-scale analysis reveals 'disruptive' innovations in research history\n\nPublished on April 1, 2026\n\nThe history of science and technology is marked by ...",{"title":27,"url":28,"summary":29,"type":21},"Scientists Unveil Innovative Method to Identify Breakthroughs in Science","https:\u002F\u002Fbioengineer.org\u002Fscientists-unveil-innovative-method-to-identify-breakthroughs-in-science\u002F","In the relentless pursuit of scientific advancement, pinpointing the precise moments that redefine knowledge remains an elusive challenge. While milestones like the theory of evolution or the inventio...",{"title":31,"url":32,"summary":33,"type":21},"New scientific fields are triggered by powerful new methods","https:\u002F\u002Fwww.nature.com\u002Farticles\u002Fs41599-025-05797-6","Abstract\n\nScientific fields embody our greatest scientific advances, but we do not yet understand how we give rise to new fields. Explaining empirically and theoretically how we kick-start new fields ...",{"title":35,"url":36,"summary":37,"type":21},"Small, Faint, or Fast, Rubin Will Find It","https:\u002F\u002Feos.org\u002Ffeatures\u002Fsmall-faint-or-fast-rubin-will-find-it","Look up at the dark night sky, and you’ll be treated to a symphony of astronomical phenomena centuries or millennia in the making. Planets make slow circuits around the Sun. Stars burn for billions of...",{"title":39,"url":40,"summary":41,"type":21},"Early Data from NSF–DOE Vera C. Rubin Observatory Reveals Over 11,000 New Asteroids","https:\u002F\u002Frubinobservatory.org\u002Fnews\u002F11000-new-asteroids","Rubin’s largest asteroid haul yet, gathered before the Legacy Survey of Space and Time even begins, is just the “tip of the iceberg.”\n\nApril 2, 2026\n\nScientists at NSF–DOE Vera C. Rubin Observatory, j...",{"title":43,"url":44,"summary":45,"type":21},"Get Involved in Rubin Research","https:\u002F\u002Frubinobservatory.org\u002Fexplore\u002Fcitizen-science","Share \n\nWhat good is a giant data set if we don't have as many eyes on it as possible, ready to make discoveries? That’s where you come in — the more people looking for particular objects or patterns ...",{"title":47,"url":48,"summary":49,"type":21},"NSF–DOE Vera C. Rubin Observatory launches real-time discovery machine for monitoring the night sky","https:\u002F\u002Fwww6.slac.stanford.edu\u002Fnews\u002F2026-02-25-nsf-doe-vera-c-rubin-observatory-launches-real-time-discovery-machine-monitoring","The NSF-DOE Vera C. Rubin Observatory, jointly funded by the U.S. National Science Foundation and the U.S. Department of Energy's Office of Science, has released its first alerts documenting astronomi...",{"title":51,"url":52,"summary":53,"type":21},"NSF-DOE Vera C. Rubin Observatory","https:\u002F\u002Fwww.nsf.gov\u002Ffocus-areas\u002Fastronomy-space\u002Frubin-observatory","Funded by the U.S. National Science Foundation and the U.S. Department of Energy's Office of Science. \n\nNSF-DOE Rubin Observatory will embark on the Legacy Survey of Space and Time, a ten-year survey ...",{"totalSources":55},9,{"generationDuration":57,"kbQueriesCount":55,"confidenceScore":58,"sourcesCount":55},116631,100,{"metaTitle":60,"metaDescription":61},"Innovative method detects scientific breakthroughs faster","Discover a scalable scan that flags research redirecting science. Explains neural-embedding maps of 55M papers\u002Fpatents and why funders care — see cases.","en","https:\u002F\u002Fimages.unsplash.com\u002Fphoto-1762028892198-3dd53a039249?ixid=M3w4OTczNDl8MHwxfHNlYXJjaHwxfHxpbm5vdmF0aXZlJTIwbWV0aG9kJTIwaWRlbnRpZnklMjBzY2llbnRpZmljfGVufDF8MHx8fDE3NzUyNzc0Mzh8MA&ixlib=rb-4.1.0&w=1200&h=630&fit=crop&crop=entropy&auto=format,compress&q=60",{"photographerName":65,"photographerUrl":66,"unsplashUrl":67},"Hossein Nasr","https:\u002F\u002Funsplash.com\u002F@nasrphotos?utm_source=coreprose&utm_medium=referral","https:\u002F\u002Funsplash.com\u002Fphotos\u002Fperson-examining-film-negatives-with-a-loupe-pneiKD4wbZI?utm_source=coreprose&utm_medium=referral",true,null,{"key":71,"name":72,"nameEn":73},"sciences","Sciences & Découvertes","Science & Discoveries",[75,77,79,81],{"text":76},"The Kojaku et al. method maps ~55 million papers and patents using dual neural embeddings to quantify “disruptiveness” as the divergence between a work’s past and future vectors.",{"text":78},"A large past–future vector gap indicates high disruption and typically corresponds to landmark discoveries or the launch of new research directions; Nobel‑level discoveries show especially large gaps.",{"text":80},"The approach detects simultaneous, distributed breakthroughs by identifying multiple papers that jointly redirect citation flows into new regions of embedding space, revealing shifts missed by citation counts.",{"text":82},"Disruptiveness complements citation metrics for funding and evaluation but is sensitive to embedding training choices, database coverage, and potential biases against underindexed languages and venues.",[84,87,90],{"question":85,"answer":86},"How does the dual‑vector embedding actually measure a “breakthrough”?","The method computes two neural embeddings for each paper or patent: a “past” vector summarizing the works it cites and a “future” vector summarizing the works that later cite it. The disruptiveness score is the magnitude of divergence between those vectors; a large divergence means subsequent research clusters away from the cited foundations and toward new directions, indicating a redirection of scientific attention. Applied to ~55 million records, the approach operationalizes breakthroughs as inflection points in idea space rather than simply high citation counts, allowing detection of works that launch new trajectories even before citation totals accumulate.",{"question":88,"answer":89},"How can funders and program officers use disruptiveness metrics responsibly?","Disruptiveness provides early signals about whether grants or programs are generating new research directions by tracking emergent embedding clusters and shifts in citation flows; funders can monitor portfolios for high‑disruption outputs to balance high‑risk, high‑reward investments against steady incremental work. Responsible use requires combining these quantitative signals with peer review, domain expertise, and replication evidence, and accounting for limitations like embedding sensitivity and coverage bias. Pilot implementations should validate disruptiveness indicators against case studies and adjust for disciplinary differences before informing major allocation decisions.",{"question":91,"answer":92},"What are the main limitations and biases of this method?","The method quantifies redirection patterns but does not prove scientific importance; neural embeddings are sensitive to model architecture, training data, and pretraining choices, which can change disruptiveness scores. Coverage gaps in citation databases can undercount work from underindexed languages, regional venues, or applied domains, producing bias against socially important but poorly indexed research. Additionally, specialized or incremental work that is practically crucial may show low disruptiveness despite high real‑world impact, so results must be interpreted alongside qualitative assessments and domain‑specific measures.",[94,102,107,113,119,124,130,136,141,145,149,154,159,164,171],{"id":95,"name":96,"type":97,"confidence":98,"wikipediaUrl":99,"slug":100,"mentionCount":101},"69d18f394eea09eba3dfe7f8","data‑intensive astronomy","concept",0.9,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FList_of_software_for_astronomy_research_and_education","69d18f394eea09eba3dfe7f8-data-intensive-astronomy",1,{"id":103,"name":104,"type":97,"confidence":98,"wikipediaUrl":105,"slug":106,"mentionCount":101},"69d18f384eea09eba3dfe7f4","impact factor","https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FImpact_factor","69d18f384eea09eba3dfe7f4-impact-factor",{"id":108,"name":109,"type":97,"confidence":110,"wikipediaUrl":111,"slug":112,"mentionCount":101},"69d18f384eea09eba3dfe7ee","neural embedding",0.96,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FWord_embedding","69d18f384eea09eba3dfe7ee-neural-embedding",{"id":114,"name":115,"type":97,"confidence":116,"wikipediaUrl":117,"slug":118,"mentionCount":101},"69d18f384eea09eba3dfe7f1","future vector",0.92,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FKRISS_Vector","69d18f384eea09eba3dfe7f1-future-vector",{"id":120,"name":121,"type":97,"confidence":122,"wikipediaUrl":69,"slug":123,"mentionCount":101},"69d18f394eea09eba3dfe7f6","simultaneous breakthroughs",0.91,"69d18f394eea09eba3dfe7f6-simultaneous-breakthroughs",{"id":125,"name":126,"type":97,"confidence":127,"wikipediaUrl":128,"slug":129,"mentionCount":101},"69d18f394eea09eba3dfe7fc","laser physics",0.88,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FLaser_science","69d18f394eea09eba3dfe7fc-laser-physics",{"id":131,"name":132,"type":97,"confidence":133,"wikipediaUrl":134,"slug":135,"mentionCount":101},"69d18f384eea09eba3dfe7f3","citation counts",0.94,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FCitation","69d18f384eea09eba3dfe7f3-citation-counts",{"id":137,"name":138,"type":97,"confidence":127,"wikipediaUrl":139,"slug":140,"mentionCount":101},"69d18f3a4eea09eba3dfe7fd","econometrics","https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FEconometrics","69d18f3a4eea09eba3dfe7fd-econometrics",{"id":142,"name":143,"type":97,"confidence":127,"wikipediaUrl":69,"slug":144,"mentionCount":101},"69d18f394eea09eba3dfe7f5","Nobel-level discoveries","69d18f394eea09eba3dfe7f5-nobel-level-discoveries",{"id":146,"name":147,"type":97,"confidence":110,"wikipediaUrl":69,"slug":148,"mentionCount":101},"69d18f384eea09eba3dfe7ef","disruptiveness (method)","69d18f384eea09eba3dfe7ef-disruptiveness-method",{"id":150,"name":151,"type":97,"confidence":116,"wikipediaUrl":152,"slug":153,"mentionCount":101},"69d18f394eea09eba3dfe7fb","randomized trials","https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FRandomized_controlled_trial","69d18f394eea09eba3dfe7fb-randomized-trials",{"id":155,"name":156,"type":97,"confidence":116,"wikipediaUrl":157,"slug":158,"mentionCount":101},"69d18f394eea09eba3dfe7fa","x-ray crystallography","https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FX-ray_crystallography","69d18f394eea09eba3dfe7fa-x-ray-crystallography",{"id":160,"name":161,"type":97,"confidence":116,"wikipediaUrl":162,"slug":163,"mentionCount":101},"69d18f384eea09eba3dfe7f0","past vector","https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FVector_WX-8","69d18f384eea09eba3dfe7f0-past-vector",{"id":165,"name":166,"type":167,"confidence":168,"wikipediaUrl":169,"slug":170,"mentionCount":101},"69d18f374eea09eba3dfe7eb","Binghamton University","organization",0.95,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FBinghamton_University","69d18f374eea09eba3dfe7eb-binghamton-university",{"id":172,"name":173,"type":167,"confidence":174,"wikipediaUrl":175,"slug":176,"mentionCount":101},"69d18f374eea09eba3dfe7ed","Science Advances",0.93,"https:\u002F\u002Fen.wikipedia.org\u002Fwiki\u002FScience_Advances","69d18f374eea09eba3dfe7ed-science-advances",[178],{"id":179,"title":180,"slug":181,"excerpt":182,"category":11,"featuredImage":183,"publishedAt":184},"69e5011994fa47eed6532e8e","OpenAI's AI models aimed at accelerating scientific discoveries","openai-s-ai-models-aimed-at-accelerating-scientific-discoveries","Introduction\n\nDrug discovery and much scientific research are slow, costly, and limited by humans’ ability to read, connect, and experimentally test ideas. Bringing a single medicine from target disco...","https:\u002F\u002Fimages.unsplash.com\u002Fphoto-1676272682018-b1435bad1cf0?ixid=M3w4OTczNDl8MHwxfHNlYXJjaHwxfHxvcGVuYWklMjBtb2RlbHMlMjBhaW1lZCUyMGFjY2VsZXJhdGluZ3xlbnwxfDB8fHwxNzc2NjE1NzA1fDA&ixlib=rb-4.1.0&w=1200&h=630&fit=crop&crop=entropy&auto=format,compress&q=60","2026-04-19T16:38:13.392Z",["Island",186],{"key":187,"params":188,"result":190},"ArticleBody_hZYBcRMyNfQGAvM0vIl7qg55EnJat5udbFhgWdFBGk",{"props":189},"{\"articleId\":\"69d18d5423cc38232fa5c574\"}",{"head":191},{}]