MATH Benchmark vs xCodeEval

Loads and preprocesses 12,500 curated competition mathematics problems from AMC 10/12, AIME, and Math Olympiads using the MATHDataset class in MATH.py. The loader supports multiple tokenization strategies and can selectively include or exclude solution steps during preprocessing, enabling researchers to evaluate models on problem-solving without solution hints. Problems are stratified across 7 mathematical subjects (Prealgebra, Algebra, Number Theory, Counting/Probability, Geometry, Intermediate Algebra, Precalculus) with structured JSON metadata including problem statements, solutions, and difficulty levels.

Unique: Curates problems exclusively from high-difficulty mathematical competitions (AMC, AIME, Olympiads) rather than generic math word problems, ensuring evaluation on reasoning-intensive problems that require multi-step derivations and deep mathematical understanding. The MATHDataset class implements subject-aware stratification enabling fine-grained evaluation across mathematical domains.

vs alternatives: More rigorous than generic math QA datasets (e.g., MathQA, SVAMP) because problems require genuine mathematical reasoning rather than simple arithmetic, making it the de facto standard for evaluating LLM mathematical capabilities in research.

Implements the is_equiv() function in math_equivalence.py that determines semantic equivalence between two mathematical expressions regardless of syntactic representation. The system applies a multi-stage normalization pipeline that handles LaTeX formatting, fraction representations, algebraic simplification, and numerical precision issues before performing string-based comparison. This enables accurate answer verification without requiring exact string matching, accommodating equivalent forms like '1/2', '0.5', and '\frac{1}{2}'.

Unique: Implements a multi-stage normalization pipeline specifically designed for competition mathematics rather than generic string comparison. The system handles domain-specific challenges like multiple valid representations of the same answer (fractions vs decimals, different LaTeX encodings) and applies algebraic simplification to catch mathematically equivalent but syntactically different forms.

vs alternatives: More robust than exact string matching or simple numerical comparison because it normalizes across multiple mathematical notations and handles algebraic equivalence, enabling accurate evaluation of LLM answers that are mathematically correct but expressed differently than ground truth.

Extracts and preserves solution steps from MATH problems, enabling evaluation of intermediate reasoning and chain-of-thought capabilities. The system can optionally include or exclude solution steps during dataset loading, supporting different evaluation methodologies: evaluating final answers only (without hints) or evaluating intermediate reasoning steps. This enables researchers to assess whether models can generate correct reasoning chains or merely guess final answers.

Unique: Preserves solution steps as first-class data throughout the evaluation pipeline, enabling evaluation of intermediate reasoning quality rather than just final answers. This supports emerging research on chain-of-thought prompting and interpretable AI reasoning.

vs alternatives: More comprehensive than final-answer-only evaluation because it assesses reasoning quality and interpretability, but requires more manual annotation and is harder to automate than simple answer verification.

Provides evaluation infrastructure in eval_math_gpt.py that runs local language models (GPT-style architectures) on MATH dataset problems with configurable inference parameters including beam search width, sampling temperature, and top-k/top-p filtering. The run_eval() function orchestrates the evaluation pipeline: loads problems from MATHDataset, generates model responses with specified decoding strategy, extracts final answers from model outputs, and compares against ground truth using mathematical equivalence checking. Supports both greedy decoding and stochastic sampling for exploring model behavior under different inference regimes.

Unique: Integrates configurable beam search and sampling directly into the evaluation loop, enabling researchers to explore how different decoding strategies affect mathematical reasoning performance. The architecture separates inference configuration from evaluation logic, allowing systematic comparison of greedy vs stochastic decoding on the same problem set.

vs alternatives: More flexible than API-based evaluation (e.g., OpenAI GPT-3 API) because it supports arbitrary inference parameters and local model variants, but requires more computational resources and manual infrastructure setup compared to cloud-based alternatives.

Provides evaluation infrastructure in evaluate_gpt3.py that interfaces with OpenAI's GPT-3 API for remote model evaluation on MATH problems. The system handles API authentication, batches problem submissions to the GPT-3 API, parses structured responses, and aggregates accuracy metrics. This enables evaluation of closed-source models without local compute resources, though with latency and cost considerations inherent to API-based inference.

Unique: Abstracts away OpenAI API complexity by providing a unified evaluation interface that handles authentication, batching, response parsing, and error handling. The system integrates seamlessly with the local evaluation pipeline, enabling side-by-side comparison of API-based and local models using identical evaluation metrics.

vs alternatives: Simpler than local evaluation for closed-source models because it eliminates infrastructure setup, but introduces API dependency, latency, and cost overhead compared to local inference on open-source models.

Aggregates evaluation results across the 12,500 problems and computes accuracy metrics stratified by mathematical subject (Prealgebra, Algebra, Number Theory, Counting/Probability, Geometry, Intermediate Algebra, Precalculus). The reporting system generates per-subject accuracy percentages, overall accuracy, and optional per-difficulty breakdowns. This enables fine-grained analysis of model strengths and weaknesses across mathematical domains, revealing whether models struggle with specific subject areas.

Unique: Implements subject-aware stratification that breaks down accuracy by mathematical domain, revealing whether models have domain-specific weaknesses (e.g., strong on Algebra but weak on Geometry). This granularity is essential for understanding model capabilities beyond aggregate accuracy.

vs alternatives: More informative than single aggregate accuracy metric because subject-stratified results expose domain-specific model limitations, enabling targeted improvement efforts and more nuanced model comparison.

Extracts and indexes structured metadata from MATH dataset JSON files including problem statement, solution steps, final answer, difficulty level, and mathematical subject. The indexing system enables efficient retrieval of problems by subject, difficulty, or other attributes, and provides structured access to problem components (problem text vs solution vs answer) for different evaluation workflows. Metadata is preserved throughout the evaluation pipeline to enable stratified analysis and filtering.

Unique: Preserves full problem metadata (subject, difficulty, solution steps) throughout the evaluation pipeline, enabling post-hoc analysis of which problem characteristics correlate with model success or failure. The indexing structure supports efficient filtering and stratified evaluation.

vs alternatives: More structured than raw problem files because metadata is parsed and indexed, enabling efficient filtering and analysis; but less flexible than custom metadata systems that could include additional annotations (e.g., required mathematical concepts, solution techniques).

Extracts final numerical or symbolic answers from model-generated text using heuristic pattern matching (e.g., regex patterns for 'Answer: X', 'Final Answer:', or boxed notation). The extraction system handles common answer formats including integers, fractions, decimals, and algebraic expressions. This enables automatic answer verification without requiring models to output structured JSON or follow strict formatting conventions, accommodating natural language model outputs.

Unique: Uses lightweight regex-based heuristics rather than requiring models to output structured JSON, enabling evaluation of base language models without answer format fine-tuning. This pragmatic approach trades robustness for flexibility, accommodating diverse model output styles.

vs alternatives: More flexible than requiring structured output because it works with any model without fine-tuning, but less reliable than models trained to output answers in standardized formats (e.g., JSON with 'answer' field).

Provides a standardized evaluation framework for code generation models that accepts generated code in 17 programming languages (C, C++, C#, Java, Kotlin, Go, Rust, Python, Ruby, PHP, JavaScript, Perl, Haskell, OCaml, Scala, D, Pascal) and validates correctness through actual execution against unit tests via the ExecEval Docker-based execution engine. Uses a centralized problem definition model with src_uid foreign keys linking generated code to shared problem descriptions and unittest_db.json, enabling consistent evaluation across language variants of the same problem.

Unique: Combines 25M training examples across 7,500 unique problems with an execution-based evaluation pipeline (ExecEval) that actually runs generated code in Docker containers against unit tests, rather than relying on static analysis or string matching. The src_uid linking system creates a normalized data model where problem descriptions and tests are stored once and referenced by all language variants, eliminating duplication and ensuring consistency.

vs alternatives: Larger scale (25M examples vs typical 10-100K) and true execution-based validation across more languages (17 vs 4-6) than HumanEval or CodeXGLUE, with explicit support for code translation and repair tasks beyond generation.

Implements a foreign key linking system where all task-specific datasets (program synthesis, code translation, APR, retrieval) reference shared problem definitions via src_uid identifiers. Problem descriptions and unit tests are stored once in centralized problem_descriptions.jsonl and unittest_db.json files, then linked by src_uid to avoid duplication. The Hugging Face datasets API automatically resolves these links during data loading, returning enriched DatasetDict objects with problem context pre-joined to task examples.

Unique: Uses a normalized relational data model (src_uid as foreign key) for a code benchmark, treating problem definitions as a separate entity layer rather than embedding them in each task dataset. This is more sophisticated than typical flat-file benchmark structures and enables consistent multi-task evaluation on identical problems.

vs alternatives: More efficient than duplicating problem descriptions across 7 task datasets (reduces storage by ~30-40%), and enables automatic link resolution via Hugging Face API unlike manual CSV joins in CodeXGLUE or HumanEval variants.

Provides a Python API for loading xCodeEval datasets from Hugging Face Hub (NTU-NLP-sg/xCodeEval) with automatic src_uid-based linking between task datasets and shared problem definitions. The datasets library handles data downloading, caching, and streaming, while the xCodeEval integration automatically joins task examples with problem_descriptions.jsonl and unittest_db.json using src_uid foreign keys. Returns DatasetDict objects with enriched examples ready for model training or evaluation.

Unique: Integrates xCodeEval with Hugging Face datasets library, providing automatic src_uid resolution and streaming support. Treats data loading as a first-class concern with built-in linking logic, rather than requiring manual JSON parsing.

vs alternatives: More convenient than manual Git LFS downloads because it handles caching and automatic linking, and integrates seamlessly with Hugging Face training pipelines vs custom data loaders.

Provides an alternative data access method using Git LFS for users who prefer direct file access or need selective dataset downloads. Supports cloning the repository with LFS disabled, then pulling specific task files or problem definitions on demand. Useful for custom processing pipelines or environments where Python/Hugging Face is not available, though requires manual src_uid linking to join task examples with problem definitions.

Unique: Provides Git LFS-based alternative to Hugging Face API, enabling direct file access and selective downloads. Requires manual src_uid linking but offers more control over data access patterns.

vs alternatives: More flexible than Hugging Face API for selective downloads and custom pipelines, but requires more manual work for src_uid linking and lacks automatic caching/streaming.

Implements a standardized three-phase evaluation pipeline (Phase 1: Generation, Phase 2: Execution, Phase 3: Metrics) that applies consistently across all 7 tasks (program synthesis, code translation, APR, tag classification, code compilation, NL-code retrieval, code-code retrieval). Phase 1 generates or retrieves code, Phase 2 executes it via ExecEval or computes retrieval metrics, and Phase 3 aggregates results into pass@k, MRR, NDCG, or other task-specific metrics. Enables direct comparison of model performance across tasks.

Unique: Defines a unified three-phase evaluation pipeline that applies to all 7 tasks, treating generation, execution, and metric computation as separate concerns. Enables consistent evaluation methodology across diverse task types (generation, translation, retrieval, classification).

vs alternatives: More comprehensive than task-specific evaluation scripts because it provides a unified framework for all 7 tasks, and enables direct comparison of model performance across different task types.

Evaluates code generation models on the program synthesis task by accepting natural language problem descriptions and generating code solutions in any of 17 languages. The evaluation pipeline (Phase 1: Generation, Phase 2: Execution, Phase 3: Metrics) runs generated code against unit tests via ExecEval, computing pass@k metrics (pass@1, pass@10, etc.) that measure the probability of finding a correct solution within k samples. Supports both single-solution and multi-sample evaluation modes for assessing model reliability.

Unique: Implements a three-phase evaluation pipeline (Generation → Execution → Metrics) with explicit pass@k computation that measures the probability of finding a correct solution within k attempts, rather than just binary pass/fail. Supports multi-sample evaluation across 17 languages with language-specific compiler configurations and timeout handling.

vs alternatives: More rigorous than HumanEval's simple pass@k because it handles language-specific compilation errors and timeouts explicitly, and scales to 25M training examples vs HumanEval's 164 problems.

Evaluates code translation models by accepting source code in one language and generated translations in a target language, then validating functional equivalence through execution against shared unit tests. The translation evaluation pipeline compiles and executes both source and translated code against the same unittest_db.json test cases, comparing outputs to detect translation errors. Supports all 17 language pairs (though not all pairs may have training data) and uses language-specific compiler mappings to handle syntax differences.

Unique: Validates code translation by executing both source and target code against identical unit tests and comparing outputs, ensuring functional equivalence rather than syntactic similarity. Uses language-specific compiler mappings to handle the complexity of 17 different compilation environments and their idiosyncrasies.

vs alternatives: More rigorous than BLEU-score-based translation metrics because it validates actual functional correctness through execution, and covers more language pairs (17 vs typical 2-4) with explicit compiler integration.

Evaluates program repair models by providing buggy code snippets and expecting corrected versions that pass unit tests. The APR evaluation pipeline executes repaired code against unittest_db.json test cases, measuring whether the repair successfully fixes the bug without introducing new failures. Supports repairs across all 17 languages and uses the same execution-based validation as program synthesis, enabling direct comparison of repair quality.

Unique: Treats program repair as an executable task where success is measured by unit test passage, rather than syntactic similarity to reference repairs. Integrates with the same ExecEval pipeline as program synthesis, enabling direct performance comparison between generation and repair models.

vs alternatives: More comprehensive than traditional APR benchmarks (Defects4J, QuixBugs) because it covers 17 languages and 7,500 problems vs 395 Java bugs, and uses consistent execution-based metrics across all repair types.