arXiv subjects

Bhanu Prakash Vangala

Publications and source records attributed to Bhanu Prakash Vangala.

9 recordsLinked to original sources

arXiv2026

FlexiFlow: Bandit-based Model Switching in ML Workflows

Model optimizations help improve inference performance and accuracy of ML workflows. However, relying on a single model to perform inference across all data batches often fails to maximize accuracy and thus overall performance. In many cases, alternate models could perform better on specific subsets of data where a primary model underperforms. Our experiments with real ML workflows indeed show that switching models improves workflow accuracy by up to 23%. Yet, current systems lack the ability to adaptively switch between models based on performance, forcing users to manually test models in sequence. We present FlexiFlow, a dataflow system that dynamically switches between alternate models when the current model exhibits low accuracy. FlexiFlow learns to rank models using a novel multi-armed bandit approach that accounts for model runtimes, probability of passing user-defined assertions, and the computational structure of the ML workflow. We show that the standard Thompson sampling approach is insufficient for switching models in ML workflows. In contrast, our proposed approaches are effective and scales to complex real-world ML workflows. Experiments show that switching models at runtime while reusing intermediate results provides higher accuracy, but also 48% efficiency gain compared to sequential workflow runs.

cs.LG↗

arXiv2026

Scores That Hold, Benchmarks That Leak: Measuring Dataset Contamination in Public Brain-Tumor MRI Classification

Automated classification of brain tumors from MRI is a heavily published application of deep learning in medical imaging, with reported accuracies on public benchmarks routinely exceeding 98%. However, accuracy does not capture a critical dimension of benchmark quality: dataset integrity, defined as the independence of test from training data at the image, patient, and acquisition-source levels. We introduce a three-layer contamination framework comprising duplicate, patient, and source-label leakage to assess the public corpora on which this literature rests. We audit the three most widely used corpora against a chest-radiograph negative control and quantify each layer's effect on measured performance across nine architectures and three evaluation conditions. Contamination is severe at every layer: 28.8% of the dominant corpus's official test split has a near-twin in its own training split, a second corpus leaks 22.3% of its test images byte-identically, 95.5% of traceable test images share a patient with training, and file-header features containing no anatomy separate tumor from no-tumor at 0.959 balanced accuracy, at parity with fine-tuned ResNet backbones. The unexpected result is that removing every identified leaked test image leaves balanced accuracy essentially unchanged: stable performance after deduplication does not establish benchmark integrity. Our findings establish dataset integrity as a distinct, measurable axis of benchmark quality that a stable leaderboard cannot certify. For biomedical research, reported accuracy on these corpora alone does not establish that a model has learned to recognize tumors rather than exploit dataset-specific cues. We release the contaminated-file lists, recovered patient identifiers, and deduplicated splits.

cs.CV↗

arXiv2026

Code That Works, Environments That Don't: Measuring Environment Reproducibility in AI-Generated Software

Code generation has emerged as a central capability of large language models, with coding agents now able to produce functionally correct software projects from natural language prompts. However, functional correctness alone does not capture a critical dimension of generation quality: environment specification, defined as the accurate identification of the dependencies required to execute generated code, is equally critical. We develop an agent protocol for environment specification and introduce a three-layer framework comprising declared, runtime-installed, and necessary-and-sufficient dependencies to systematically assess coding agents for environment specification. Using this protocol, we evaluate the extent to which coding agents systematically misspecify software environment dependencies and how this misspecification varies across three agents, four languages, and fifty programming tasks. Our results show that current coding agents exhibit systematic generalization failures along this dimension, producing dependency specifications that are inconsistent, redundant, or incomplete in ways that functional tests do not detect. Across agents, dependency set agreement is as low as 7% for identical tasks, and newer agents show no meaningful improvement, suggesting the failure is not resolved by scale or recency. The largest divergence occurs between the declared and runtime dependency layers, implicating environment priors learned from the models' training distributions as the primary driver. Our findings establish environment specification as a distinct, measurable axis of code generation quality that current benchmarks do not capture, and motivate training objectives and evaluation protocols that jointly optimize for functional correctness and environmental portability.

cs.SE↗

arXiv2026

Evaluation Choices Shape Biomedical ML Claims: A Pediatric Pneumonia Benchmark Case Study

Biomedical machine learning papers often compress model performance into one headline number. That number can look like a property of the model even when it depends strongly on how the benchmark was evaluated. We study this problem on the widely used Kermany pediatric chest radiograph dataset using nine image classifiers and a controlled evaluation protocol. Under the same protocol, the eight pretrained backbones differ by only 0.026 AUROC. In contrast, changing whether the backbone is frozen or fine-tuned changes AUROC by 0.044 on average, and changing the decision threshold changes balanced accuracy by 0.090 on average. The official test split is also measurably different from the training pool: a partition classifier distinguishes them at AUC 0.697, rising to 0.898 for normal radiographs. Most strikingly, a classifier using only file properties, with no image anatomy, reaches 0.992 balanced accuracy within the training pool but falls to 0.496 on the official test split. Validation-fitted thresholds and calibration also transfer imperfectly. These results show that a high benchmark score can support different conclusions when the split, training policy, threshold, metric, calibration, and uncertainty are not communicated with it. We end with a seven-item reporting recommendation in which each item is tied to an effect measured in the study

cs.CV↗

arXiv2026

One Threshold Does Not Fit All Languages: Language-Conditional Deferral for Reliable and Efficient Low-Resource Text Classification

In the Global South, the lower-income countries of Africa, Asia, and Latin America where most of the world's languages are spoken, a deployed text classifier usually runs on ordinary CPUs, serves many languages with a single model, has few labeled examples in any of them, and relies on people to catch its mistakes. Such a system is only useful if it can promise how often it will be wrong: at most a fixed fraction of the labels it assigns on its own may be incorrect, and everything else must go to a person. Split conformal prediction delivers this promise through a single confidence threshold, normally estimated on validation data pooled across languages. We ask whether the promise reaches every language, and it does not. On MasakhaNEWS (16 African languages) and AfriSenti (12 languages plus two never seen in training), a pooled threshold meets the 90% target on average but covers Somali at 77.5%, Tigrinya at 83.7%, and the two unseen languages at 77.5% and 81.2%. Estimating one threshold per language brings every language to between 89.1% and 91.0% without retraining, and it shows how unequal the cost of the promise is: keeping it means sending 43% of Somali news and over 80% of Amharic and Xitsonga tweets to a person, against under 8% of Nigerian Pidgin news. One or two hundred labels per language are enough and the models train in minutes on one CPU core, so the fix is affordable: calibrate, report, and budget human review one language at a time.

cs.CL↗

arXiv2026

How Many Labels Does a Language Need? Annotation Budgets and Cross-Lingual Pooling for African-Language Text Classification

Every text classifier for an African language begins with a budgeting question: how many labelled examples are needed, and can labels from other African languages stand in for them? We answer both questions empirically for 28 language-task pairs, news topic classification in 16 languages (MasakhaNEWS) and tweet sentiment in 12 languages (AfriSenti), using a character n-gram linear model that trains in seconds on two CPU cores with no pretrained weights and no accelerator. Monolingual learning curves at budgets from 25 to several thousand labels show that topic classification reaches 90\% of its full-data macro-F1 with about 400 labels in the median language, while sentiment is still improving at the full training size in 11 of 12 languages and needs thousands of labels. Pooling the full training data of the other languages in the benchmark is worth a great deal at small budgets and nothing at large ones: at 25 target labels it adds 0.20 macro-F1 on average for news (up to 0.43 for Lingala) and 0.08 for sentiment, the gain decays to zero by 800 labels, and at full size pooling hurts in 9 of 16 and 8 of 12 languages. Twenty-five target labels plus pooled data match what 100 to 400 monolingual labels achieve for most news languages. A complete zero-shot transfer matrix shows that transfer without any target labels recovers a median of only 13\% (news) and 4\% (sentiment) of the gap between a majority-class predictor and the in-language model, with the exceptions explained by shared script (Amharic and Tigrinya), shared lexicon (English and Nigerian Pidgin, the Arabic dialects), or a shared label prior rather than by language family. We release code that regenerates every number from the public benchmark files and translate the results into concrete annotation guidance for teams building African-language classifiers without GPUs.

cs.CL↗

arXiv2026

AI-Generated Code Is Not Reproducible (Yet): An Empirical Study of Dependency Gaps in LLM-Based Coding Agents

The rise of Large Language Models (LLMs) as coding agents promises to accelerate software development, but their impact on generated code reproducibility remains largely unexplored. This paper presents an empirical study investigating whether LLM-generated code can be executed successfully in a clean environment with only OS packages and using only the dependencies that the model specifies. We evaluate three state-of-the-art LLM coding agents (Claude Code, OpenAI Codex, and Gemini) across 300 projects generated from 100 standardized prompts in Python, JavaScript, and Java. We introduce a three-layer dependency framework (distinguishing between claimed, working, and runtime dependencies) to quantify execution reproducibility. Our results show that only 68.3% of projects execute out-of-the-box, with substantial variation across languages (Python 89.2%, Java 44.0%). We also find a 13.5 times average expansion from declared to actual runtime dependencies, revealing significant hidden dependencies.

cs.SE↗

arXiv2025

HalluMat: Detecting Hallucinations in LLM-Generated Materials Science Content Through Multi-Stage Verification

Artificial Intelligence (AI), particularly Large Language Models (LLMs), is transforming scientific discovery, enabling rapid knowledge generation and hypothesis formulation. However, a critical challenge is hallucination, where LLMs generate factually incorrect or misleading information, compromising research integrity. To address this, we introduce HalluMatData, a benchmark dataset for evaluating hallucination detection methods, factual consistency, and response robustness in AI-generated materials science content. Alongside this, we propose HalluMatDetector, a multi-stage hallucination detection framework that integrates intrinsic verification, multi-source retrieval, contradiction graph analysis, and metric-based assessment to detect and mitigate LLM hallucinations. Our findings reveal that hallucination levels vary significantly across materials science subdomains, with high-entropy queries exhibiting greater factual inconsistencies. By utilizing HalluMatDetector verification pipeline, we reduce hallucination rates by 30% compared to standard LLM outputs. Furthermore, we introduce the Paraphrased Hallucination Consistency Score (PHCS) to quantify inconsistencies in LLM responses across semantically equivalent queries, offering deeper insights into model reliability.

cs.AI↗

arXiv2025

Efficient Multi-Model Orchestration for Self-Hosted Large Language Models

Self-hosting large language models (LLMs) is increasingly appealing for organizations seeking privacy, cost control, and customization. Yet deploying and maintaining in-house models poses challenges in GPU utilization, workload routing, and reliability. We introduce Pick and Spin, a practical framework that makes self-hosted LLM orchestration scalable and economical. Built on Kubernetes, it integrates a unified Helm-based deployment system, adaptive scale-to-zero automation, and a hybrid routing module that balances cost, latency, and accuracy using both keyword heuristics and a lightweight DistilBERT classifier. We evaluate four models, Llama-3 (90B), Gemma-3 (27B), Qwen-3 (235B), and DeepSeek-R1 (685B) across eight public benchmark datasets, with five inference strategies, and two routing variants encompassing 31,019 prompts and 163,720 inference runs. Pick and Spin achieves up to 21.6% higher success rates, 30% lower latency, and 33% lower GPU cost per query compared with static deployments of the same models.

cs.DC↗