ARXIV:2602.12413 · DATA CONTAMINATION · SUBMITTED 19 MAR · 18:48 UTC · FRESHNESS STALE

VerifiedSource: PDF linkedPartialPaperPack: 3 of 4 citation fields filledMissingMissing fields: authorsPartialProof: unverified proof status

Soft Contamination Means Benchmarks Test Shallow Generalization

arXiv

A method to detect soft contamination in LLM training data affecting benchmark performance.

Blocked on Code›Score3.0Evidence unverified

Opportunity summary

Pain A method to detect soft contamination in LLM training data affecting benchmark performance.

Evidence 0 refs | 0 sources | 33% coverage

Blocker Evidence unverified

Open Build Read PDF Signal Canvas Track

PROBLEM

A method to detect soft contamination in LLM training data affecting benchmark performance. Typical decontamination filters use n-gram matching which fail to detect semantic duplicates: sentences with equivalent (or near-equivalent) content that are not…

METHOD

Full abstract

If LLM training data is polluted with benchmark test data, then benchmark performance gives biased estimates of out-of-distribution (OOD) generalization. Typical decontamination filters use n-gram matching which fail to detect semantic duplicates: sentences with equivalent (or near-equivalent) content that are not close in string space. We study this soft contamination of training data by semantic duplicates. Among other experiments, we embed the Olmo3 training corpus and find that: 1) contamination remains widespread, e.g. we find semantic duplicates for 78% of CodeForces and exact duplicates for 50% of ZebraLogic problems; 2) including semantic duplicates of benchmark data in training does improve benchmark performance; and 3) when finetuning on duplicates of benchmark datapoints, performance also improves on truly-held-out datapoints from the same benchmark. We argue that recent benchmark gains are thus confounded: the prevalence of soft contamination means gains reflect both genuine capability improvements and the accumulation of test data and effective test data in growing training corpora.

RESULT

ScienceToStartup currently rates this 3.0/10 on the public viability pass. we find semantic duplicates for 78% of CodeForces and exact duplicates for 50% of ZebraLogic problems; 2) including semantic duplicates of benchmark data in…

WHY NOW

Data Contamination moved forward this cycle; last verified April 2026. Public score 3.0/10.

Continue into Read for claims, analysis, references, and neighboring papers.

Opportunity summary

Score3.0

PainA method to detect soft contamination in LLM training data affecting benchmark performance.

Evidence0 refs | 0 sources | 33% coverage

Blockermissing authors

Analysis summary

A method to detect soft contamination in LLM training data affecting benchmark performance.

VerifiedSource: PDF linkedPartialPaperPack: 3 of 4 citation fields filledMissingMissing fields: authorsPartialProof: unverified proof status

ARXIV:2602.12413 · DATA CONTAMINATION · SUBMITTED 19 MAR · 18:48 UTC · FRESHNESS STALE

VerifiedSource: PDF linkedPartialPaperPack: 3 of 4 citation fields filledMissingMissing fields: authorsPartialProof: unverified proof status

Soft Contamination Means Benchmarks Test Shallow Generalization

arXiv

A method to detect soft contamination in LLM training data affecting benchmark performance.

Blocked on Code›Score3.0Evidence unverified

Opportunity summary

Pain A method to detect soft contamination in LLM training data affecting benchmark performance.

Evidence 0 refs | 0 sources | 33% coverage

Blocker Evidence unverified

Open Build Read PDF Signal Canvas Track

PROBLEM

METHOD

Full abstract

RESULT

WHY NOW

Data Contamination moved forward this cycle; last verified April 2026. Public score 3.0/10.

Continue into Read for claims, analysis, references, and neighboring papers.

Opportunity summary

Score3.0

PainA method to detect soft contamination in LLM training data affecting benchmark performance.

Evidence0 refs | 0 sources | 33% coverage

Blockermissing authors

Analysis summary

A method to detect soft contamination in LLM training data affecting benchmark performance.

VerifiedSource: PDF linkedPartialPaperPack: 3 of 4 citation fields filledMissingMissing fields: authorsPartialProof: unverified proof status

Paper Pack

10.48550/arXiv.2602.12413

Soft Contamination Means Benchmarks Test Shallow Generalization

A method to detect soft contamination in LLM training data affecting benchmark performance.

Abstract

Source availability

PDF linked

The paper record includes a public PDF URL.

Extraction status

Derived fallback

Read summaries are estimated from adjacent metadata, not verified extraction rows.

Proof status

unverified

0 refs; 0 sources; 33% coverage.

What was readable

linkedon filenot materializedderived fallback28 indexednot indexed

Derived fallback: Estimated from adjacent evidence; not verified from source.

Viability

3.0

Time to MVP

MVP estimate missing

Commercial

No commercial flags on file

Export

Preparing verified analysis

lens / founder

PROBLEM

METHOD

RESULT

WHY NOW

Data Contamination moved forward this cycle; last verified April 2026. Public score 3.0/10.

Claim map

Abstract-backed public claims while anchored extraction refreshes.

Strong 0Mixed 0Weak 4

Evidencepartial
A method to detect soft contamination in LLM training data affecting benchmark performance. Typical decontamination filters use n-gram matching which fail to detect semantic duplicates: sentences with equivalent (or near-equivalent) content that are not close in string space.
Implicationpartial
Abstract-backed fallback claim; anchored extraction has not materialized a public claim row yet.
Verificationpartial
partial
Evidencepartial
If LLM training data is polluted with benchmark test data, then benchmark performance gives biased estimates of out-of-distribution (OOD) generalization. Typical decontamination filters use n-gram matching which fail to detect semantic duplicates: sentences with equivalent (or near-equivalent) content that are not close in string space.
Implicationpartial
Abstract-backed fallback claim; anchored extraction has not materialized a public claim row yet.
Verificationpartial
partial
Evidencepartial
ScienceToStartup currently rates this 3.0/10 on the public viability pass. we find semantic duplicates for 78% of CodeForces and exact duplicates for 50% of ZebraLogic problems; 2) including semantic duplicates of benchmark data in training does improve benchmark performance; and 3) when finetuning on duplicates of benchmark datapoints, performance also improves on truly-held-out datapoints from the same benchmark.
Implicationpartial
Abstract-backed fallback claim; anchored extraction has not materialized a public claim row yet.
Verificationpartial
partial
Evidencepartial
Data Contamination moved forward this cycle; last verified April 2026. Public score 3.0/10.
Implicationpartial
Abstract-backed fallback claim; anchored extraction has not materialized a public claim row yet.
Verificationpartial
partial

Constellation map

Paper-native neighborhood for concepts, methods, materials, markets, and competitors. Missing lanes stay labeled instead of disappearing behind commercialization gates.

Open full Signal Canvas

Concepts

not indexed

Methods

Materials

PDF linked

Markets

Data Contamination

Competitors

not indexed

Competitive landscape

A method to detect soft contamination in LLM training data affecting benchmark performance.

Segment

Data Contamination

Adoption evidence

No public code link in the paper record yet

Commercial read

3.0/10 public viability

Direct

not classified

Adjacent

not classified

Substitute

not classified

Unknown

not classified

Buzz

No indexed public discussion is attached to 2602.12413 yet. That is a visibility signal, not a blank module: the monitor is watching the public channels below.

Hacker News

Not indexed yet

Bluesky

Not indexed yet

PDF

Preview the source document here, or use the hero PDF action for a new tab.

References(28)

Llama-Embed-Nemotron-8B: A Universal Text Embedding Model for Multilingual and Cross-Lingual Tasks

2025Yauhen Babakhin, Radek Osmulski et al.

LessLeak-Bench: A First Investigation of Data Leakage in LLMs Across 83 Software Engineering Benchmarks

2025Xin Zhou, M. Weyssow et al.

ZebraLogic: On the Scaling Limits of LLMs for Logical Reasoning

2025Bill Yuchen Lin, Ronan Le Bras et al.

Overestimation in LLM Evaluation: A Controlled Large-Scale Study on Data Contamination's Impact on Machine Translation

2025Muhammed Yusuf Kocyigit, Eleftheria Briakou et al.

SSA: Semantic Contamination of LLM-Driven Fake News Detection

2025Cheng Xu, Nan Yan et al.

The mosaic memory of large language models.

2024Igor Shilov, Matthieu Meeus et al.

Quantifying Contamination in Evaluating Code Generation Capabilities of Language Models

2024Martin Riddell, Ansong Ni et al.

Rephrasing the Web: A Recipe for Compute and Data-Efficient Language Modeling

2024Pratyush Maini, Skyler Seto et al.

Investigating Data Contamination for Pre-training Language Models

2024Minhao Jiang, Ken Ziyu Liu et al.

Rethinking Benchmark and Contamination for Language Models with Rephrased Samples

2023Shuo Yang, Wei-Lin Chiang et al.

What's In My Big Data?

2023Yanai Elazar, Akshita Bhagia et al.

Detecting Pretraining Data from Large Language Models

2023Weijia Shi, Anirudh Ajith et al.

MuSR: Testing the Limits of Chain-of-thought with Multistep Soft Reasoning

2023Zayne Sprague, Xi Ye et al.

True Detective: A Deep Abductive Reasoning Benchmark Undoable for GPT-3 and Challenging for GPT-4

2022Maksym Del, Mark Fishel

Self-Instruct: Aligning Language Models with Self-Generated Instructions

2022Yizhong Wang, Yeganeh Kordi et al.

MTEB: Massive Text Embedding Benchmark

2022Niklas Muennighoff, Nouamane Tazi et al.

Data Contamination: From Memorization to Exploitation

2022Inbal Magar, Roy Schwartz

Chain of Thought Prompting Elicits Reasoning in Large Language Models

2022Jason Wei, Xuezhi Wang et al.

Training Verifiers to Solve Math Word Problems

2021K. Cobbe, Vineet Kosaraju et al.

Program Synthesis with Large Language Models

2021Jacob Austin, Augustus Odena et al.

Showing 20 of 28 references

CITED BY

No citing papers are indexed in the public S2S graph yet. This is an explicit zero-signal state, not a hidden lookup.

Foundation

Prior WorkThe Reliability Gap in Benchmark Auditing: Distribution Shift and Scale as Failure Modes of Contamination Detection

3.0

Extension

none indexed

Commercially relevant

Higher ViabilitySilicon Bureaucracy and AI Test-Oriented Education: Contamination Sensitivity and Score Confidence in LLM Benchmarks

4.0

Higher ViabilityNo Memorization, No Detection: Output Distribution-Based Contamination Detection in Small Language Models

5.0

Higher ViabilityCross-Context Verification: Hierarchical Detection of Benchmark Contamination through Session-Isolated Analysis

7.0

Higher ViabilityAre Large Language Models Truly Smarter Than Humans?

4.0

Higher ViabilityNavigating the Sea of LLM Evaluation: Investigating Bias in Toxicity Benchmarks

6.0

Higher ViabilitySALMUBench: A Benchmark for Sensitive Association-Level Multimodal Unlearning

7.0

Higher ViabilityData filtering methods for training language models

4.0

Higher ViabilityIn-the-Wild Model Organisms: Mitigating Undesirable Emergent Behaviors in Production LLM Post-Training via Data Attribution

5.0

Higher ViabilityBeyond the Class Subspace: Teacher-Guided Training for Reliable Out-of-Distribution Detection in Single-Domain Models

5.0

Conflicting

none indexed

Owned Distribution

Subscribe to the weekly brief

Get the weekly shortlist of commercializable papers, benchmark movers, and proof receipts that matter for product execution.

Agent drawer

5 surfaces preserved for agents. Humans can ignore.

Developer contracts, payload previews, evidence maps, and run controls stay here instead of the Read, Build, and Track workspace.

Run context

Paper: 2602.12413
Route: /paper/soft-contamination-means-benchmarks-test-shallow-generalization
Active tab: read
Artifact: soft-contamination-means-benchmarks-test-shallow-generalization

Available agents

Read extractor
Build planner
Track monitor
Competitive mapper
Related-paper scout

API/MCP endpoints

REST paper pack API/api/v1/paper/soft-contamination-means-benchmarks-test-shallow-generalization/paper-pack
REST build passport API/api/v1/paper/soft-contamination-means-benchmarks-test-shallow-generalization/build-passport
REST OpenAPI/api/openapi.json
MCP descriptor/api/mcp
MCP resourcesciencetostartup://surfaces/paper-workspace

Tool contracts

paper_packbuild_passportopportunity_kernelforesightsource_proofevidence_state

Payload preview

Inspect payload

{
  "contract_version": "paper-r2",
  "paper_id": "20d656bc-05f9-450f-bcda-0b423d841c1d",
  "arxiv_id": "2602.12413",
  "canonical_route": "/paper/soft-contamination-means-benchmarks-test-shallow-generalization",
  "active_tab": "synced from current hash by the drawer client",
  "selected_artifact": "soft-contamination-means-benchmarks-test-shallow-generalization",
  "endpoints": {
    "paper_pack": "/api/v1/paper/soft-contamination-means-benchmarks-test-shallow-generalization/paper-pack",
    "build_passport": "/api/v1/paper/soft-contamination-means-benchmarks-test-shallow-generalization/build-passport",
    "mcp_resource": "sciencetostartup://surfaces/paper-workspace"
  }
}

Schema validation

paper-r2 contract: present
JSON-LD twin: SSR emitted
OpenAPI path parity: /api/openapi.json
MCP resource parity: paper-workspace

Job trace

queued: drawer opened by user action
running: inspect or copy payload
succeeded: payload available in SSR
failed: route errors appear in evidence cards

Evidence map

sources used: page freshness, source proof anchors, JSON-LD
missing sources: exposed by PaperPack and EvidenceState chips
derived fallbacks: marked unverified before handoff

Page Freshness

Canonical route, proof status, last verified, refs, sources, and coverage.

Page Freshness

Paper proof surface

Canonical route: /paper/soft-contamination-means-benchmarks-test-shallow-generalization

stale

Proof freshness: stale
Proof status: unverified
Display score: 3/10
Last proof check: 2026-03-19
Score updated: 2026-04-02
Score fresh until: 2026-05-02
References: 0
Source count: 0
Coverage: 33%

This page is showing the last landed evidence receipt and score bundle because the latest proof data is outside the freshness window.

OpenAlex: pending — this preprint is not yet indexed by OpenAlex.

Agent Handoff

Endpoint list, payload shape, route context, and copyable handoff data.

Agent Handoff

Soft Contamination Means Benchmarks Test Shallow Generalization

Canonical ID soft-contamination-means-benchmarks-test-shallow-generalization | Route /paper/soft-contamination-means-benchmarks-test-shallow-generalization

REST example

curl https://sciencetostartup.com/api/v1/agent-handoff/paper/soft-contamination-means-benchmarks-test-shallow-generalization

MCP example

{
  "tool": "get_paper",
  "arguments": {
    "arxiv_id": "2602.12413"
  }
}

source_context

{
  "surface": "paper",
  "mode": "paper",
  "query": "Soft Contamination Means Benchmarks Test Shallow Generalization",
  "normalized_query": "2602.12413",
  "route": "/paper/soft-contamination-means-benchmarks-test-shallow-generalization",
  "paper_ref": "soft-contamination-means-benchmarks-test-shallow-generalization",
  "topic_slug": null,
  "benchmark_ref": null,
  "dataset_ref": null
}

Buildability Receipt

Verdict, compute envelope, blockers, signature state, and receipt links.

Paper proof page receipt window

Not build-ready: Soft Contamination Means Benchmarks Test Shallow Generalization

/buildability/soft-contamination-means-benchmarks-test-shallow-generalization

Ignoreblocked

Subject: Soft Contamination Means Benchmarks Test Shallow Generalization

Verdict

Ignore

Verdict is Ignore because current viability and proof state do not clear the buildability gate.

Time to first demo

Insufficient data

No first-demo timestamp, owner estimate, or elapsed demo receipt is attached to this surface.

Compute envelope

Structured compute envelope

Insufficient data

No data, compute, hardware, memory, latency, dependency, or serving requirement receipt is attached.

Evidence ids

Receipt path

/buildability/soft-contamination-means-benchmarks-test-shallow-generalization

Paper ref

soft-contamination-means-benchmarks-test-shallow-generalization

arXiv id

2602.12413

Freshness

Generated at

2026-03-19T18:48:05.835Z

Evidence freshness

stale

Last verification

2026-03-19T18:48:05.835Z

Sources

References

Coverage

33%

Hash state

Lineage hash

7439b8c4a4b90b855e6122f036e13cfe46fab15b65c71f240a7416684ddf2782

Canonical opportunity-kernel lineage hash.

Signature state

External signature

unsigned_external

No founder, registry, pilot, or production-adoption signature is attached to this receipt.

Verification

not_verified

Verification is blocked until an external signature is provided.

Blockers

Missing: repo_url
Missing: references
Missing: distribution_readiness_scores
Missing: paper_extraction_scorecards
Unknown: distribution readiness has not been computed yet

Verification pending / evidence receipt incomplete

repo_url

references

Missing proof, requirement, signature, approval, adoption, or telemetry fields are blockers and must not be inferred.

Open receipt API receipt Build Loop Signal Canvas Proof divergence Divergence API Brier outcomes API

Source Proof anchors

Visual citations from the paper document graph.

JSON-LD twin

The application/ld+json payload rendered for agents.

{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "WebPage",
      "@id": "https://sciencetostartup.com/paper/soft-contamination-means-benchmarks-test-shallow-generalization#webpage",
      "url": "https://sciencetostartup.com/paper/soft-contamination-means-benchmarks-test-shallow-generalization",
      "name": "Soft Contamination Means Benchmarks Test Shallow Generalization",
      "description": "A method to detect soft contamination in LLM training data affecting benchmark performance.",
      "isPartOf": {
        "@id": "https://sciencetostartup.com/#website"
      }
    },
    {
      "@type": "ScholarlyArticle",
      "@id": "https://sciencetostartup.com/paper/soft-contamination-means-benchmarks-test-shallow-generalization#scholarlyArticle",
      "headline": "Soft Contamination Means Benchmarks Test Shallow Generalization",
      "description": "A method to detect soft contamination in LLM training data affecting benchmark performance.",
      "url": "https://sciencetostartup.com/paper/soft-contamination-means-benchmarks-test-shallow-generalization",
      "sameAs": "https://arxiv.org/abs/2602.12413",
      "identifier": {
        "@type": "PropertyValue",
        "propertyID": "arXiv",
        "value": "2602.12413"
      },
      "isAccessibleForFree": true,
      "isPartOf": {
        "@id": "https://sciencetostartup.com/#website"
      },
      "datePublished": "2026-02-12T21:12:53.000Z",
      "citation": [
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "a2a819feab6d7863afe67372a2ff8813d4538dfc"
          },
          "url": "https://www.semanticscholar.org/paper/a2a819feab6d7863afe67372a2ff8813d4538dfc"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "0684edb230b65d2cd7ce63f50b72ab8606c12ee1"
          },
          "url": "https://www.semanticscholar.org/paper/0684edb230b65d2cd7ce63f50b72ab8606c12ee1"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "ea2da86fd9d6488a4935b6c135658accc8f64d6c"
          },
          "url": "https://www.semanticscholar.org/paper/ea2da86fd9d6488a4935b6c135658accc8f64d6c"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "b6846ea6dc98f86e6cad8080f462f5d03276a890"
          },
          "url": "https://www.semanticscholar.org/paper/b6846ea6dc98f86e6cad8080f462f5d03276a890"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "7269d290675e4bf4edb87224d23705f1359287ba"
          },
          "url": "https://www.semanticscholar.org/paper/7269d290675e4bf4edb87224d23705f1359287ba"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "6bf28ebbb8df92582bc53a7fe49016e0caa4c074"
          },
          "url": "https://www.semanticscholar.org/paper/6bf28ebbb8df92582bc53a7fe49016e0caa4c074"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "2905dc5ad70b462f4f5543df3047dffadb5c0e4e"
          },
          "url": "https://www.semanticscholar.org/paper/2905dc5ad70b462f4f5543df3047dffadb5c0e4e"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "b2fda33b7c122c044a7faa185d250d59ce9e4453"
          },
          "url": "https://www.semanticscholar.org/paper/b2fda33b7c122c044a7faa185d250d59ce9e4453"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "227b5f8206b64858edeef6723b96af14133077e3"
          },
          "url": "https://www.semanticscholar.org/paper/227b5f8206b64858edeef6723b96af14133077e3"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "8a583a8bee70fa59186664d0df5afb60a1591a21"
          },
          "url": "https://www.semanticscholar.org/paper/8a583a8bee70fa59186664d0df5afb60a1591a21"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "3422d5e0cdfdc935d6a84a1e3d3f96659265fe3a"
          },
          "url": "https://www.semanticscholar.org/paper/3422d5e0cdfdc935d6a84a1e3d3f96659265fe3a"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "743ef29a9406c44c835684c7755d423d6ca0b663"
          },
          "url": "https://www.semanticscholar.org/paper/743ef29a9406c44c835684c7755d423d6ca0b663"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "256ef1f8d0ea2982cc50d3e85e5f1b4920f037fe"
          },
          "url": "https://www.semanticscholar.org/paper/256ef1f8d0ea2982cc50d3e85e5f1b4920f037fe"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "e65b346d442e9962a4276dc1c1af2956d9d5f1eb"
          },
          "url": "https://www.semanticscholar.org/paper/e65b346d442e9962a4276dc1c1af2956d9d5f1eb"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "88a74e972898de887ad9587d4c87c3a9f03f1dc5"
          },
          "url": "https://www.semanticscholar.org/paper/88a74e972898de887ad9587d4c87c3a9f03f1dc5"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "fd6ade42ad3e0b9ffe418a37c4cb4b16346d90b7"
          },
          "url": "https://www.semanticscholar.org/paper/fd6ade42ad3e0b9ffe418a37c4cb4b16346d90b7"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "1b6e810ce0afd0dd093f789d2b2742d047e316d5"
          },
          "url": "https://www.semanticscholar.org/paper/1b6e810ce0afd0dd093f789d2b2742d047e316d5"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "d6045d2ccc9c09ca1671348de86d07da6bc28eea"
          },
          "url": "https://www.semanticscholar.org/paper/d6045d2ccc9c09ca1671348de86d07da6bc28eea"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "a38e0f993e4805ba8a9beae4c275c91ffcec01df"
          },
          "url": "https://www.semanticscholar.org/paper/a38e0f993e4805ba8a9beae4c275c91ffcec01df"
        },
        {
          "@type": "ScholarlyArticle",
          "identifier": {
            "@type": "PropertyValue",
            "propertyID": "SemanticScholar",
            "value": "db1afe3b3cd4cd90e41fbba65d3075dd5aebb61e"
          },
          "url": "https://www.semanticscholar.org/paper/db1afe3b3cd4cd90e41fbba65d3075dd5aebb61e"
        }
      ],
      "additionalProperty": [
        {
          "@type": "PropertyValue",
          "propertyID": "viabilityScore",
          "value": 3
        },
        {
          "@type": "PropertyValue",
          "propertyID": "researchDomain",
          "value": "Data Contamination"
        }
      ]
    },
    {
      "@type": "BreadcrumbList",
      "itemListElement": [
        {
          "@type": "ListItem",
          "position": 1,
          "name": "Home",
          "item": "https://sciencetostartup.com"
        },
        {
          "@type": "ListItem",
          "position": 2,
          "name": "Data Contamination",
          "item": "https://sciencetostartup.com/topics"
        },
        {
          "@type": "ListItem",
          "position": 3,
          "name": "Soft Contamination Means Benchmarks Test Shallow Generalizat",
          "item": "https://sciencetostartup.com/paper/soft-contamination-means-benchmarks-test-shallow-generalization"
        }
      ]
    }
  ]
}

Soft Contamination Means Benchmarks Test Shallow Generalization

Soft Contamination Means Benchmarks Test Shallow Generalization

Claim map

Constellation map

Competitive landscape

Buzz

PDF

References(28)

Related Papers

Subscribe to the weekly brief

Build artifacts

Brief

Experiment plan

Validation checklist

Scientific founder

Translational engineer

Domain operator

GTM lead

Regulatory/clinical advisor

Timeline

Claim map

Constellation map

Competitive landscape

Buzz

PDF

References(28)

Related Papers

Subscribe to the weekly brief

Build artifacts

Brief

Experiment plan

Validation checklist

Scientific founder

Translational engineer

Domain operator

GTM lead

Regulatory/clinical advisor

Timeline