Negative / Null Result ReportOpen accessComputer Science
Rylan Malarchick · 2026 · arXiv
We present an analysis of quantum circuit fidelity across the full compilation stack, from high-level gate optimization through pulse-level control. We connect a C++ circuit optimizer to a per-gate Lindblad master-equation fidelity model whose decoherence channels are cross-validated against qiskit-dynamics and whose absolute predictions are benchmarked against execution on real hardware. Across a campaign of 4,452 experiment runs over 371 benchmark circuits, gate cancellation provides the dominant improvement ($d = 1.66$, 72% of circuits improved), while circuit size and pulse duration are th
Negative / Null Result ReportOpen accessComputer Science
Song Tae-Eun · 2026 · arXiv
Cross-Context Review (CCR) improves LLM verification by separating production and review into independent sessions. A natural extension is multi-turn review: letting the reviewer ask follow-up questions, receive author responses, and review again. We call this Dynamic Cross-Context Review (D-CCR). In a controlled experiment with 30 artifacts and 150 injected errors, we tested four D-CCR variants against the single-pass CCR baseline. Single-pass CCR (F1 = 0.376) significantly outperformed all multi-turn variants, including D-CCR-2b with question-and-answer exchange (F1 = 0.303, $p < 0.001$, $d
Negative / Null Result ReportOpen accessComputer Science
Marco Cognetta, Tatsuya Hiraoka, Naoaki Okazaki et al. · 2024 · arXiv
We explore threshold vocabulary trimming in Byte-Pair Encoding subword tokenization, a postprocessing step that replaces rare subwords with their component subwords. The technique is available in popular tokenization libraries but has not been subjected to rigorous scientific scrutiny. While the removal of rare subwords is suggested as best practice in machine translation implementations, both as a means to reduce model size and for improving model performance through robustness, our experiments indicate that, across a large space of hyperparameter settings, vocabulary trimming fails to improv
Failed Experiment ReportOpen accessComputer Science
Frits de Nijs, Daan Wilmer · 2012 · arXiv
The goal of this paper is to critically evaluate a heuristic algorithm for the Inverse Banzhaf Index problem by Laruelle and Widgrén. Few qualitative results are known about the approximation quality of the heuristics for this problem. The intuition behind the operation of this approximation algorithm is analysed and evaluated. We found that the algorithm can not handle general inputs well, and often fails to improve inputs. It is also shown to diverge after only tens of iterations. We present three alternative extensions of the algorithm that do not alter the complexity but can result in up t
Failed Experiment ReportOpen accessPhysics
Nicolas Tancogne-Dejean, Martin Lüders, Carsten A. Ullrich · 2023 · arXiv
We extend some of the well established self-interaction correction (SIC) schemes of density-functional theory to the case of systems with noncollinear magnetism. Our proposed SIC schemes are tested on a set of molecules and metallic clusters in combination with the widely used local spin-density approximation. As expected from the collinear SIC, we show that the averaged-density SIC works well for improving ionization energies but fails to improve more subtle quantities like the dipole moments of polar molecules. We investigate the exchange-correlation magnetic field produced by our extension
Negative / Null Result ReportOpen accessComputer Science
Janarthanan Rajendran, Jonathan K. Kummerfeld, Satinder Singh · 2021 · arXiv
For each goal-oriented dialog task of interest, large amounts of data need to be collected for end-to-end learning of a neural dialog system. Collecting that data is a costly and time-consuming process. Instead, we show that we can use only a small amount of data, supplemented with data from a related dialog task. Naively learning from related data fails to improve performance as the related data can be inconsistent with the target task. We describe a meta-learning based method that selectively learns from the related dialog task data. Our approach leads to significant accuracy improvements in
Negative / Null Result ReportOpen accessComputer Science
Dhruvi Khandelwal, Anurag Basistha, Ayushi Jolotia et al. · 2026 · arXiv
Deep learning proxies for Alternating Current Optimal Power Flow (ACOPF) lack systematic methods for determining architectural size. This paper conducts a constructive thought experiment to answer a fundamental inquiry: how wide must a neural network be to almost accurately approximate the ACOPF manifold? We introduce a Loss-Guided Neural Densification (LG-ND) algorithm that incrementally discovers necessary capacity by expanding only when the current deep neural network topology fails to improve further. Empirical results across various IEEE systems show that LG-ND achieves performance parity
Negative / Null Result ReportOpen accessComputer Science
Runzhe Zhan, Xuebo Liu, Derek F. Wong et al. · 2021 · arXiv
Meta-learning has been sufficiently validated to be beneficial for low-resource neural machine translation (NMT). However, we find that meta-trained NMT fails to improve the translation performance of the domain unseen at the meta-training stage. In this paper, we aim to alleviate this issue by proposing a novel meta-curriculum learning for domain adaptation in NMT. During meta-training, the NMT first learns the similar curricula from each domain to avoid falling into a bad local optimum early, and finally learns the curricula of individualities to improve the model robustness for learning dom
Negative / Null Result ReportOpen accessComputer Science
Arijit Sehanobish, Charles Lovering · 2026 · arXiv
We study the \textit{parameter placement problem}: given a fixed budget of $k$ trainable entries within the B matrix of a LoRA adapter (A frozen), does the choice of which $k$ matter? Under supervised fine-tuning, random and informed subsets achieve comparable performance. Under GRPO on base models, random placement fails to improve over the base model, while gradient-informed placement recovers standard LoRA accuracy. This regime dependence traces to gradient structure: SFT gradients are low-rank and directionally stable, so any subset accumulates coherent updates; GRPO gradients are high-ran
Negative / Null Result ReportOpen accessComputer Science
Zelu Deng, Yujie Zhong, Sheng Guo et al. · 2021 · arXiv
This work aims at improving instance retrieval with self-supervision. We find that fine-tuning using the recently developed self-supervised (SSL) learning methods, such as SimCLR and MoCo, fails to improve the performance of instance retrieval. In this work, we identify that the learnt representations for instance retrieval should be invariant to large variations in viewpoint and background etc., whereas self-augmented positives applied by the current SSL methods can not provide strong enough signals for learning robust instance-level representations. To overcome this problem, we propose InsCL
Negative / Null Result ReportOpen accessMathematics
Oliver J. Hines, Christian L. Hines · 2026 · arXiv
Gradient boosted decision trees require a stopping rule to avoid overfitting. The standard rule monitors a validation loss and stops if the loss fails to improve for a fixed patience period. However, the patience parameter has no interpretable scale and validation losses can be noisy or implicitly defined by a user-specified gradient. We propose ScoreStop, a gradient-based early-stopping rule that casts the stopping decision at each iteration as a test of the null hypothesis that the current predictor is the population risk minimizer. We use a functional score test, computed on validation data
Negative / Null Result ReportOpen accessComputer Science
Hanbing Liu, Haoyang Li, Xiaokang Zhang et al. · 2025 · arXiv
Direct Preference Optimization (DPO) has proven effective in complex reasoning tasks like math word problems and code generation. However, when applied to Text-to-SQL datasets, it often fails to improve performance and can even degrade it. Our investigation reveals the root cause: unlike math and code tasks, which naturally integrate Chain-of-Thought (CoT) reasoning with DPO, Text-to-SQL datasets typically include only final answers (gold SQL queries) without detailed CoT solutions. By augmenting Text-to-SQL datasets with synthetic CoT solutions, we achieve, for the first time, consistent and
Negative / Null Result ReportOpen accessComputer Science
Àlex Miranda-Pascual, Javier Parra-Arnau, Thorsten Strufe · 2026 · arXiv
Sampling is renowned for its privacy amplification in differential privacy (DP), and is often assumed to improve the utility of a DP mechanism by allowing a noise reduction. In this paper, we further show that this last assumption is flawed: When measuring utility at equal privacy levels, sampling as preprocessing consistently yields penalties due to utility loss from omitting records over all canonical DP mechanisms -- Laplace, Gaussian, exponential, and report noisy max -- , as well as recent applications of sampling, such as clustering. Extending this analysis, we investigate suppression as
Negative / Null Result ReportOpen accessEngineering
Haoyang Li, Yuchen Hu, Chen Chen et al. · 2024 · arXiv
Deep neural network (DNN)-based speech enhancement (SE) usually uses conventional activation functions, which lack the expressiveness to capture complex multiscale structures needed for high-fidelity SE. Group-Rational KAN (GR-KAN), a variant of Kolmogorov-Arnold Networks (KAN), retains KAN's expressiveness while improving scalability on complex tasks. We adapt GR-KAN to existing DNN-based SE by replacing dense layers with GR-KAN layers in the time-frequency (T-F) domain MP-SENet and adapting GR-KAN's activations into the 1D CNN layers in the time-domain Demucs. Results on Voicebank-DEMAND sho
Failed Experiment ReportOpen accessComputer Science
Zixian Huang, Kaichen Yang, Xu Huang et al. · 2026 · arXiv
A widely adopted strategy for model enhancement is to use synthetic data generated by a stronger model for supervised fine-tuning (SFT). However, for emerging reasoning models like Qwen3-8B, this approach often fails to improve reasoning capabilities and can even lead to a substantial drop in performance. In this work, we identify substantial stylistic divergence between teacher generated data and the distribution of student as a major factor impacting SFT. To bridge this gap, we propose a Teacher-Student Cooperation Data Synthesis framework (TESSY), which interleaves teacher and student model
Negative / Null Result ReportOpen accessComputer Science
Piyawat Lertvittayakumjorn, David Kinney, Vinodkumar Prabhakaran et al. · 2025 · arXiv
Generative large language models (LLMs) have demonstrated gaps in diverse cultural awareness across the globe. We investigate the effect of retrieval augmented generation and search-grounding techniques on LLMs' ability to display familiarity with various national cultures. Specifically, we compare the performance of standard LLMs, LLMs augmented with retrievals from a bespoke knowledge base (i.e., KB grounding), and LLMs augmented with retrievals from a web search (i.e., search grounding) on multiple cultural awareness benchmarks. We find that search grounding significantly improves the LLM p
Negative / Null Result ReportOpen accessComputer Science
Fırat Öncel, Matthias Bethge, Beyza Ermis et al. · 2024 · arXiv
In the last decade, the generalization and adaptation abilities of deep learning models were typically evaluated on fixed training and test distributions. Contrary to traditional deep learning, large language models (LLMs) are (i) even more overparameterized, (ii) trained on unlabeled text corpora curated from the Internet with minimal human intervention, and (iii) trained in an online fashion. These stark contrasts prevent researchers from transferring lessons learned on model generalization and adaptation in deep learning contexts to LLMs. To this end, our short paper introduces empirical ob
Negative / Null Result ReportOpen accessComputer Science
Adam Byerly, Daniel Khashabi · 2024 · arXiv
Self-consistency (SC) improves the performance of large language models (LLMs) across various tasks and domains that involve short content. However, does this support its effectiveness for long-context problems? We challenge the assumption that SC's benefits generalize to long-context settings, where LLMs often struggle with position bias, the systematic over-reliance on specific context regions-which hinders their ability to utilize information effectively from all parts of their context. Through comprehensive experimentation with varying state-of-the-art models, tasks, and SC formulations, w
Negative / Null Result ReportOpen accessComputer Science
Chunliang Li, Tianze Cao, Sanyuan Zhao · 2026 · arXiv
Visual Autoregressive (VAR) modeling inefficiently applies a fixed computational depth to each position when generating high-resolution images. While existing methods accelerate inference by pruning tokens using frequency maps, their binary hard-pruning approach is fundamentally limited and fails to improve quality even with better frequency estimation. Observing that VAR models possess significant depth redundancy, we propose a paradigm shift from pruning entire tokens to adaptively allocating per-token computational depth. To this end, we introduce DepthVAR, a training-free framework that dy
Negative / Null Result ReportOpen accessComputer Science
Amanpreet Singh, Mike D'Arcy, Arman Cohan et al. · 2022 · arXiv
Learned representations of scientific documents can serve as valuable input features for downstream tasks without further fine-tuning. However, existing benchmarks for evaluating these representations fail to capture the diversity of relevant tasks. In response, we introduce SciRepEval, the first comprehensive benchmark for training and evaluating scientific document representations. It includes 24 challenging and realistic tasks, 8 of which are new, across four formats: classification, regression, ranking and search. We then use this benchmark to study and improve the generalization ability o
Negative / Null Result ReportOpen accessComputer Science
Ruth Cohen, Lu Feng, Ayala Bloch et al. · 2026 · arXiv
While natural-language explanations from large language models (LLMs) are widely adopted to improve transparency and trust, their impact on objective human-AI team performance remains poorly understood. We identify a Persuasion Paradox: fluent explanations systematically increase user confidence and reliance on AI without reliably improving, and in some cases undermining, task accuracy. Across three controlled human-subject studies spanning abstract visual reasoning (RAVEN matrices) and deductive logical reasoning (LSAT problems), we disentangle the effects of AI predictions and explanations u
Negative / Null Result ReportOpen accessComputer Science
Andrei Liviu Nicolicioiu, Mohammad Pezeshki, Aaron Courville · 2026 · arXiv
On-policy self-distillation achieves strong pass@1 accuracy by using a single model as both teacher and student, with the teacher conditioned on a correct demonstration to provide dense token-level feedback. We show that this could come at a hidden cost: rollout diversity decreases and pass@k curves flatten (i.e., generating more rollouts fails to improve accuracy). We trace this to compounding biases in the design of self-distillation with sampled demonstrations. The teacher scores each student rollout while conditioned on a sampled correct rollout, channeling its feedback through the model's
Negative / Null Result ReportOpen accessComputer Science
OFM Riaz Rahman Aranya, Kevin Desai · 2026 · arXiv
Synthetic data, an appealing alternative to extensive expert-annotated data for medical image segmentation, consistently fails to improve segmentation performance despite its visual realism. The reason being that synthetic and real medical images exist in different semantic feature spaces, creating a domain gap that current semi-supervised learning methods cannot bridge. We propose SRA-Seg, a framework explicitly designed to align synthetic and real feature distributions for medical image segmentation. SRA-Seg introduces a similarity-alignment (SA) loss using frozen DINOv2 embeddings to pull s
Negative / Null Result ReportOpen accessEngineering
Sreyan Ghosh, Mohammad Sadegh Rasooli, Michael Levit et al. · 2024 · arXiv
Generative Error Correction (GEC) has emerged as a powerful post-processing method to enhance the performance of Automatic Speech Recognition (ASR) systems. However, we show that GEC models struggle to generalize beyond the specific types of errors encountered during training, limiting their ability to correct new, unseen errors at test time, particularly in out-of-domain (OOD) scenarios. This phenomenon amplifies with named entities (NEs), where, in addition to insufficient contextual information or knowledge about the NEs, novel NEs keep emerging. To address these issues, we propose DARAG (D
Negative / Null Result ReportMedicine
Yang, Mueller, D'Andrea et al. · 2026 · Academic psychiatry : the journal of the American Association of Directors of Psychiatric Residency Training and the Association for Academic Psychiatry
Resident physicians experience high rates of depression, anxiety, burnout, and loneliness, yet few evidence-based interventions have been evaluated in this population. This randomized controlled pilot trial examined the feasibility,…
View details →DOI: 10.1007/s40596-026-02384-y Negative / Null Result ReportOpen accessComputer Science
Julian Coda-Forno, Zhuokai Zhao, Qiang Zhang et al. · 2025 · arXiv
Should LLM reasoning live in a separate module, or within a single model's forward pass and representational space? We study dual-architecture latent reasoning, where a fluent Base exchanges latent messages with a Coprocessor, and test two hypotheses aimed at improving latent communication over Liu et al. (2024): (H1) increase channel capacity; (H2) learn communication via joint finetuning. Under matched latent-token budgets on GPT-2 and Qwen-3, H2 is consistently strongest while H1 yields modest gains. A unified soft-embedding baseline, a single model with the same forward pass and shared rep
Negative / Null Result ReportOpen accessComputer Science
Romain Cosentino, Sarath Shekkizhar, Adam Earle et al. · 2026 · arXiv
Negotiation requires more than inferring what the other side wants: it requires using that information to make advantageous offers and counteroffers over multiple turns. We study whether large language model (LLM) agents do this in a controlled multi-attribute bargaining environment. We find that current LLM agents can model a counterparty's preferences, but do not reliably turn that knowledge into strategic bargaining. When given negotiating partner preference information, agents model it accurately and early in their reasoning traces, yet this does not reliably improve outcomes for the infor
Negative / Null Result ReportOpen accessComputer Science
Mike Zhang, Ali Basirat, Desmond Elliott · 2026 · arXiv
Prior work establishes that controlled contrastiveness between self-generated responses from large language models, set via reward scores, improves downstream preference tuning in English. We extend this method to multiple languages and evaluate two models across a total of 14 high and low-resource languages on a diverse set of tasks. Our central finding is that cross-lingual contrastive preference tuning on self-generations (CroCo) transfers without language-specific preference annotation. A reward model trained on English preferences (atop a multilingual base) produces useful within-language
Negative / Null Result ReportOpen accessComputer Science
Zhiwei Jia, Xuanlin Li, Zhan Ling et al. · 2022 · arXiv
Generalization in deep reinforcement learning over unseen environment variations usually requires policy learning over a large set of diverse training variations. We empirically observe that an agent trained on many variations (a generalist) tends to learn faster at the beginning, yet its performance plateaus at a less optimal level for a long time. In contrast, an agent trained only on a few variations (a specialist) can often achieve high returns under a limited computational budget. To have the best of both worlds, we propose a novel generalist-specialist training framework. Specifically, w
Negative / Null Result ReportOpen accessComputer Science
Lily H. Zhang, Rajesh Ranganath · 2023 · arXiv
Methods which utilize the outputs or feature representations of predictive models have emerged as promising approaches for out-of-distribution (OOD) detection of image inputs. However, these methods struggle to detect OOD inputs that share nuisance values (e.g. background) with in-distribution inputs. The detection of shared-nuisance out-of-distribution (SN-OOD) inputs is particularly relevant in real-world applications, as anomalies and in-distribution inputs tend to be captured in the same settings during deployment. In this work, we provide a possible explanation for SN-OOD detection failur
