The paper examines whether transcripts of model solutions can reveal capabilities after benchmark scores saturate. By simulating saturation with historical data, the authors test protocols that use only correctly solved transcripts to predict performance on unsaturated items. Pairwise judging of transcripts outperforms other methods, reducing prediction error by 5.3×, though its effectiveness can vary across distribution shifts.
arXiv:2605.30018v3 Announce Type: replace-cross Abstract: Large language models (LLMs) frequently achieve impressive scores on standardized benchmarks, yet accuracy alone offers a limited view of the...
The paper introduces SuperValid, a framework that generates out-of-distribution, capability-aligned validation data by distilling core concepts from benchmarks and expanding them into diverse, knowledge-rich texts. By focusing on capability-level performance rather than benchmark-specific metrics, SuperValid’s loss correlates strongly and stably with downstream benchmark results across a wide range of models, scales, and training data distributions. This training‑free metric can be computed during training, enabling model selection, early stopping, and scaling decisions without the need for benchmark evaluation.
arXiv:2607. 28801v1 Announce Type: cross Abstract: Benchmark datasets are central to evaluating Large Language Models (LLMs), yet they are typically conceived as monolithic tasks, obscuring substantial variation in the demands of individual samples.
arXiv:2602. 15327v2 Announce Type: replace-cross Abstract: Machine learning model performance improvements tend to arise from competition and application.
arXiv:2605.24614v2 Announce Type: replace-cross Abstract: Large language model (LLM) unlearning has emerged as a crucial post-hoc mechanism for privacy protection and AI safety, yet auditing whether...
arXiv:2607. 20465v1 Announce Type: new Abstract: The quality of training data fundamentally determines the capabilities of large language models (LLMs), yet no unified benchmark exists to measure how well LLMs, agents, and data-centric workflows actually prepare training data end to end.
arXiv:2602. 14307v4 Announce Type: replace Abstract: As frontier Large Language Models (LLMs) increasingly saturate new benchmarks shortly after they are published, benchmarking itself is at a juncture: if frontier models keep improving, it will become increasingly hard for humans to generate discriminative tasks, provide accurate ground-truth answers, or evaluate complex solutions.
arXiv:2607. 28008v2 Announce Type: replace-cross Abstract: Representation engineering reads and steers capability directions in large language models, yet methods are typically evaluated on paper-specific synthetic data.
arXiv:2606. 26836v1 Announce Type: new Abstract: Existing benchmarks typically report accuracy for a single model on a single run.
arXiv:2607. 02032v1 Announce Type: new Abstract: Evaluating LLM agents on benchmarks like SWE-Bench and GAIA can be expensive, time-consuming, and requires complex infrastructure.
The paper introduces SITAR, a situational distortion metric for evaluating automatic speech recognition (ASR) systems. SITAR aggregates situational severity and cumulative extent into a unified score using a bounded recursive formulation, and is implemented as a lightweight BERT-family regressor that requires minimal runtime cost. The authors validate SITAR through extrinsic and intrinsic protocols, showing it tracks capability scaling and provides diagnostic utility across diverse datasets and models.
arXiv:2607. 25891v1 Announce Type: new Abstract: Evaluating AI agents in interactive environments is hindered by fragmented tasks, scaffolds, verifiers, and scoring rules.