Grading Needs a Rubric, Not Intelligence

Small language models can grade open-ended examination answers as reliably as substantially more expensive models when they grade against an explicit rubric. We test this claim as the design principle behind any-to-bench: a frontier model reads source documents once, at ingestion, to extract each question and its rubric; lower-cost models then perform all repeated grading work. We evaluate six cost-efficient model configurations from two model families at three reasoning-effort levels. Each configuration answers 24 open-ended examination questions, and each also grades every answer sheet three times, yielding 3,456 per-question grades. Scores depend overwhelmingly on the answer being graded: answer identity explains 95.6% of score variance, whereas judge identity explains only 0.2%. Raising a writer's reasoning effort moves earned scores by as much as 0.143 of full marks, while raising a judge's reasoning effort moves assigned scores by at most 0.006. Six frontier-tier judges, added as a check, reproduce these scores and are no more reliable as a panel. Two ablations then decompose the rubric on the same questions and answers. Removing its criteria and levels while keeping the official answer changes nothing measurable. Removing the official answer as well collapses reliability (ICC 0.888 to 0.628), inflates scores, and makes judge reasoning effort matter again. The rubric is what decouples grading from judge intelligence, and within the rubric the official answer does nearly all the work. We find no evidence of length preference or same-family preference under rubric-anchored grading.

ChartGenEval: Corruption-Tested Multi-Dimensional Feedback for Rhythm-Game Chart Generation

A generated rhythm-game chart need not reproduce one official note sequence: many note choices can fit the same song and difficulty. Reference-note agreement therefore measures reconstruction, not the full design problem. We introduce ChartGenEval, a six-question evaluation framework with an automatic, corruption-tested core. It leaves note choice open while anchoring timing to the song: the matched official chart supplies only its authored timing map, never target notes. We test each core output with dose-controlled failures rather than assume that a familiar statistic measures chart quality. Across 80 held-out song groups, seven output axes satisfy prespecified sensitivity and invariance criteria in nine nonredundant tests. Complementary stress tests on the 40-song development panel expose two broader lessons. A chart-wide phase estimate recovers injected shifts of 15, 30, and 60 ms while chart-only outputs remain essentially unchanged. Common-pattern rewriting lowers mean language-model perplexity by 37%, and loop collapse raises mean self-similarity by 62%. ChartGenEval therefore reports separate, role-specific signals instead of one proxy or total score. This profile provides automatic feedback for comparing and iterating generators; selected outputs are candidate optimization targets or constraints after task-specific stress testing.

BUILD-AND-FIND: An Effort-Aware Protocol for Evaluating Agent-Managed Codebases

Most coding-agent benchmarks ask whether generated code behaves correctly. That remains essential, but repository-level engineering is increasingly agent-managed: one agent writes a repository, and later agents inspect, audit, or extend it as working context. In that setting, a generated repository is not only an answer to a task but also a communication artifact for future work. Even when strong agents nearly satisfy the visible behavioral objective, repositories can differ in how clearly they expose the intended behavior and design choices behind that behavior. We introduce BUILD-AND-FIND, a protocol for evaluating whether downstream agents can recover those intended choices from generated repositories, and how much inspection that recovery requires. For each task, a builder sees a hidden repository specification and creates a codebase; a finder sees only the codebase and a specification-traced multiple-choice question bank. The protocol separates behavioral correctness from artifact-side recovery and reports recovery accuracy, repeatability, implementation coverage, and inspection effort. Accuracy and stability act as gates: effort is interpreted only when recovery succeeds reliably. Among artifacts from which the same intent can be recovered, lower effort by the same finder suggests that the artifact makes that intent easier to locate. Question-only and spec-only controls quantify generic priors and specification access, while audits separate omitted claims from finder failures and check whether correct answers cite artifact evidence. In the released high-prior task pack, recovery accuracy is near saturation, so inspection effort and finder-specific effects provide the main panel-local comparison.

Session-Level Spoken Language Assessment with a Multimodal Foundation Model via Multi-Target Learning

Spoken Language Assessment (SLA) estimates a learner's oral proficiency from spontaneous speech. The growing population of L2 English speakers has intensified the demand for reliable SLA, a critical component of Computer Assisted Language Learning (CALL). Existing efforts often rely on cascaded pipelines, which are prone to error propagation, or end-to-end models that often operate on a short audio window, which might miss discourse-level evidence. This paper introduces a novel multimodal foundation model approach that performs session-level evaluation in a single pass. Our approach couples multi-target learning with a frozen, Whisper ASR model-based speech prior for acoustic-aware calibration, allowing for jointly learning holistic and trait-level objectives of SLA without resorting to handcrafted features. By coherently processing the entire response session of an L2 speaker, the model excels at predicting holistic oral proficiency. Experiments conducted on the Speak & Improve benchmark demonstrate that our proposed approach outperforms the previous state-of-the-art cascaded system and exhibits robust cross-part generalization, producing a compact deployable grader that is tailored for CALL applications.

The NTNU System at the S&I Challenge 2025 SLA Open Track

A recent line of research on spoken language assessment (SLA) employs neural models such as BERT and wav2vec 2.0 (W2V) to evaluate speaking proficiency across linguistic and acoustic modalities. Although both models effectively capture features relevant to oral competence, each exhibits modality-specific limitations. BERT-based methods rely on ASR transcripts, which often fail to capture prosodic and phonetic cues for SLA. In contrast, W2V-based methods excel at modeling acoustic features but lack semantic interpretability. To overcome these limitations, we propose a system that integrates W2V with Phi-4 multimodal large language model (MLLM) through a score fusion strategy. The proposed system achieves a root mean square error (RMSE) of 0.375 on the official test set of the Speak & Improve Challenge 2025, securing second place in the competition. For comparison, the RMSEs of the top-ranked, third-ranked, and official baseline systems are 0.364, 0.384, and 0.444, respectively.

A Novel Data Augmentation Approach for Automatic Speaking Assessment on Opinion Expressions

Automated speaking assessment (ASA) on opinion expressions is often hampered by the scarcity of labeled recordings, which restricts prompt diversity and undermines scoring reliability. To address this challenge, we propose a novel training paradigm that leverages a large language models (LLM) to generate diverse responses of a given proficiency level, converts responses into synthesized speech via speaker-aware text-to-speech synthesis, and employs a dynamic importance loss to adaptively reweight training instances based on feature distribution differences between synthesized and real speech. Subsequently, a multimodal large language model integrates aligned textual features with speech signals to predict proficiency scores directly. Experiments conducted on the LTTC dataset show that our approach outperforms methods relying on real data or conventional augmentation, effectively mitigating low-resource constraints and enabling ASA on opinion expressions with cross-modal information.

Acoustically Precise Hesitation Tagging Is Essential for End-to-End Verbatim Transcription Systems

Verbatim transcription for automatic speaking assessment demands accurate capture of disfluencies, crucial for downstream tasks like error analysis and feedback. However, many ASR systems discard or generalize hesitations, losing important acoustic details. We fine-tune Whisper models on the Speak & Improve 2025 corpus using low-rank adaptation (LoRA), without recourse to external audio training data. We compare three annotation schemes: removing hesitations (Pure), generic tags (Rich), and acoustically precise fillers inferred by Gemini 2.0 Flash from existing audio-transcript pairs (Extra). Our challenge system achieved 6.47% WER (Pure) and 5.81% WER (Extra). Post-challenge experiments reveal that fine-tuning Whisper Large V3 Turbo with the "Extra" scheme yielded a 5.5% WER, an 11.3% relative improvement over the "Pure" scheme (6.2% WER). This demonstrates that explicit, realistic filled-pause labeling significantly enhances ASR accuracy for verbatim L2 speech transcription.