GDP.pdf Benchmark Leaderboard
Publication
View on arXivGDP.pdf: A Benchmark for Professional Document Reasoning
GDP.pdf All-pass
Score
GDP.pdf: All-pass
Domain Results
GDP.pdf: Mean Pass by Domain
Cost
Costs include contestant model calls only. Judge calls and local PDF preparation and OCR are excluded.
GDP.pdf: Cost per Task
Token Usage
GDP.pdf: Output Tokens per Task
Speed
Time per task estimates contestant response-generation time from output-token usage and model output speed. It excludes judge calls and local PDF preparation and OCR.
GDP.pdf: Time per Task
Example Tasks
Explore five representative GDP.pdf tasks, their source documents, reviewed model submissions, and criterion-level judge rationales.
Instructions
According to the attached file, which EDS subtype(s) is associated with TNXB gene mutations and what are the differentiating characteristics of the subtype(s)?
Reference document
Provided to the model
Submission result
One reviewed model submission
Model submission
Gemini 3.5 Flash
Based on the provided research review, mutations in the TNXB gene (which encodes the tenascin-X protein) are associated with the following Ehlers-Danlos syndrome (EDS) subtypes and clinical presentations:
1. Associated EDS Subtypes
- Classical-like EDS (clEDS): Caused by autosomal recessive, biallelic mutations (homozygous or compound heterozygous) in TNXB, leading to a complete deficiency of tenascin-X (TNX).
- Contiguous Gene Syndrome (CAH-X): Caused by a contiguous gene deletion or mutation overlapping both the TNXB gene and the adjacent CYP21A2 (previously CYP21B) gene.
- TNXB Haploinsufficiency: Associated with heterozygous state mutations in TNXB that result in reduced (but not completely absent) serum TNX levels, which clinically overlaps with hypermobile EDS (hEDS).
2. Differentiating Characteristics
Classical-like EDS (clEDS)
- Primary Features: Marked by generalized joint hypermobility (GJH), soft/velvety and hyperextensible skin, and easy bruising.
- Key Differentiating Characteristics:
- Unlike classical EDS, individuals with clEDS lack the typical atrophic (papyraceous) scarring.
- It follows an autosomal recessive pattern of inheritance, whereas classical EDS is typically autosomal dominant.
CAH-X (Contiguous Gene Syndrome)
- Primary Features: Combines the connective tissue features of EDS with the endocrine features of congenital adrenal hyperplasia (CAH) due to 21-hydroxylase deficiency.
- Key Differentiating Characteristics:
- Patients with CAH-X exhibit significantly higher frequencies of joint hypermobility, chronic joint pain, joint dislocations, and piezogenic papules on the feet compared to patients with CAH alone.
- They are also predisposed to structural cardiac valve abnormalities.
TNXB Haploinsufficiency
- Primary Features: Present in heterozygous carriers (such as obligate family members of clEDS patients) who have roughly half of the normal level of serum TNX.
- Key Differentiating Characteristics:
- Individuals present with joint hypermobility, chronic joint pain, and joint subluxations.
- Unlike those with complete tenascin-X deficiency (clEDS), skin hyperextensibility and easy bruising are generally absent in these heterozygous individuals.
Criterion verdicts
Independent judge results and concise rationales
The answer correctly identifies classical-like EDS (clEDS) as caused by biallelic TNXB mutations leading to complete tenascin-X deficiency.
The answer correctly states that classical-like EDS results from biallelic TNXB mutations leading to a complete deficiency of tenascin-X (TNX) protein.
The answer correctly identifies TNX-deficient clEDS as caused by autosomal-recessive biallelic TNXB variants, explicitly including homozygous and compound heterozygous mutations. This satisfies the stated genetic-basis criterion.
The answer identifies generalized joint hypermobility as a primary feature of classical-like EDS (clEDS), which satisfies the criterion's required common feature.
The response explicitly lists soft/velvety skin as a primary feature of classical-like EDS (clEDS), satisfying the criterion.
The answer lists hyperextensible skin as a primary feature of classical-like EDS (clEDS), which satisfies the criterion’s required common feature.
The response explicitly lists easy bruising as a primary feature of classical-like EDS (clEDS), satisfying the criterion for a common cEDS feature.
The answer explicitly identifies TNXB haploinsufficiency as associated with heterozygous TNXB variants and states that it clinically overlaps with hypermobile EDS (hEDS), including hEDS-like features such as joint hypermobility, chronic pain, and subluxations.
The answer explicitly states that TNX-deficient classical-like EDS lacks the typical atrophic (papyraceous) scarring seen in classical EDS, directly satisfying the criterion.
The answer explicitly states that TNX-deficient classical-like EDS (clEDS) is caused by biallelic mutations and follows an autosomal recessive inheritance pattern.
The answer explicitly lists “Contiguous Gene Syndrome (CAH-X)” as a separate associated EDS subtype, contrary to the criterion requiring that CAH-X not be treated as a distinct EDS subtype.
Explore Evaluations
Artificial Analysis Intelligence Index v4.2A composite benchmark aggregating ten challenging evaluations to provide a holistic measure of AI capabilities across mathematics, science, coding, and reasoning.
Artificial Analysis Openness IndexA composite measure providing an industry standard to communicate model openness for users and developers.
AA-Briefcase: Agentic Knowledge Work BenchmarkA private evaluation developed by Artificial Analysis for frontier agentic capability in long-horizon knowledge work, testing agents on realistic business workflows that require deliverables such as spreadsheets, presentations, and memos.
GDPval-AA v2 LeaderboardGDPval-AA v2 is Artificial Analysis' evaluation framework for OpenAI's GDPval dataset. It tests AI models on real-world tasks across 44 occupations and 9 major industries. Models are given shell access and web browsing capabilities in an agentic loop via Stirrup to solve tasks, with Elo ratings derived from blind pairwise comparisons.
APEX-Agents-AA Benchmark LeaderboardArtificial Analysis' implementation of the APEX-Agents benchmark, testing AI agents on long-horizon, cross-application tasks in professional-services environments with realistic application tooling.
AA-AnalystAgent Benchmark LeaderboardArtificial Analysis' data analysis benchmark, testing AI agents on their ability to work with spreadsheets and documents to answer quantitative questions a Business Analyst or Data Analyst would face day-to-day.
AutomationBench-AA: Agentic SaaS Workflow BenchmarkA benchmark measuring agentic task completion across simulated SaaS application environments, scoring the share of each task's objectives completed without guardrail violations.
Harvey LAB-AA Benchmark LeaderboardArtificial Analysis' implementation of Harvey's Legal Agent Benchmark (LAB), testing AI agents on real-world legal work from Harvey's dataset of 120 private tasks spanning 24 legal practice areas. The agent reads case documents in a sandbox and produces legal deliverables (e.g., memos, disclosure schedules, deposition summaries), graded criterion-by-criterion by a single LLM rubric judge.
GDP.pdf Benchmark LeaderboardArtificial Analysis' implementation of Surge AI's GDP.pdf benchmark, testing whether language models can reason over long, real-world professional documents and satisfy detailed task-specific criteria.
EnterpriseOps-Gym-AA Benchmark LeaderboardArtificial Analysis' independent implementation of ServiceNow's EnterpriseOps-Gym, an agentic benchmark testing whether LLM agents can complete stateful, multi-step enterprise workflows across eight business domains via live tool use, graded on the final state of the underlying databases.
𝜏³-Banking Benchmark LeaderboardA fintech customer-support benchmark from the 𝜏-Knowledge framework that tests whether agents can navigate a large unstructured knowledge base and execute multi-step tool calls to resolve realistic banking workflows.
Terminal-Bench v2.1 Benchmark LeaderboardA verified refresh of Terminal-Bench v2.0 — 89 curated tasks across software engineering, system administration, data processing, model training, and security, with environment and instruction fixes so scores reflect agent capability rather than environment gaps.
Artificial Analysis Long Context Reasoning Benchmark LeaderboardA challenging benchmark measuring language models' ability to extract, reason about, and synthesize information from long-form documents ranging from 10k to 100k tokens (measured using the cl100k_base tokenizer).
AA-Omniscience: Knowledge and Hallucination BenchmarkA benchmark measuring factual recall and hallucination across various economically relevant domains.
SciCode Benchmark LeaderboardA scientist-curated coding benchmark featuring 288 test set subproblems from 80 laboratory problems across 16 scientific disciplines.
Humanity's Last Exam Benchmark LeaderboardA frontier-level benchmark with 2,500 expert-vetted questions across mathematics, sciences, and humanities, designed to be the final closed-ended academic evaluation.
CritPt Benchmark LeaderboardA benchmark designed to test LLMs on research-level physics reasoning tasks, featuring 71 composite research challenges.
The most challenging 198 questions from GPQA, where PhD experts achieve 65% accuracy but skilled non-experts only reach 34% despite web access.
ITBench-AA Benchmark LeaderboardArtificial Analysis' implementation of IBM's ITBench benchmark, testing AI agents on Kubernetes incident root-cause analysis from offline incident snapshots. The agent inspects alerts, events, traces, and topology and identifies the contributing-factor entities (deployments, pods, namespaces, network policies, etc.) responsible for the failure.
MMMU-Pro Benchmark LeaderboardAn enhanced MMMU benchmark that eliminates shortcuts and guessing strategies to more rigorously test multimodal models across 30 academic disciplines.
IFBench Benchmark LeaderboardA benchmark evaluating precise instruction-following generalization on 58 diverse, verifiable out-of-domain constraints that test models' ability to follow specific output requirements.
Terminal-Bench Hard Benchmark LeaderboardAn agentic benchmark evaluating AI capabilities in terminal environments through software engineering, system administration, and data processing tasks.
𝜏²-Bench Telecom Benchmark LeaderboardA dual-control conversational AI benchmark simulating technical support scenarios where both agent and user must coordinate actions to resolve telecom service issues.
MMLU-Pro Benchmark LeaderboardAn enhanced version of MMLU with 12,000 graduate-level questions across 14 subject areas, featuring ten answer options and deeper reasoning requirements.
LiveCodeBench Benchmark LeaderboardA contamination-free coding benchmark that continuously harvests fresh competitive programming problems from LeetCode, AtCoder, and CodeForces, evaluating code generation, self-repair, and execution.
MATH-500 Benchmark LeaderboardA 500-problem subset from the MATH dataset, featuring competition-level mathematics across six domains including algebra, geometry, and number theory.
AIME 2025 Benchmark LeaderboardAll 30 problems from the 2025 American Invitational Mathematics Examination, testing olympiad-level mathematical reasoning with integer answers from 000-999.
Global-MMLU-Lite Benchmark LeaderboardA lightweight, multilingual version of MMLU, designed to evaluate knowledge and reasoning skills across a diverse range of languages and cultural contexts.
Medical Long Context Reasoning (MLCR-AA)An open benchmark from Wisedocs measuring how well models reason over long, fragmented medical records, performing the multi-document synthesis claims professionals rely on when reviewing insurance and healthcare cases.