AINeutralarXiv – CS AI · Jun 115/10
🧠SemantiClean is a modular framework that extracts semantic signals from e-commerce session data to predict purchase intent and customer behavior while prioritizing auditability and reproducibility over raw predictive accuracy. The system uses a predefined library of 24 behavioral elements organized across four layers and implements safeguards against signal inflation, representing a shift toward transparent, governance-focused AI systems over conventional black-box optimizers.
AINeutralarXiv – CS AI · Jun 116/10
🧠Researchers demonstrate that sparsified Kolmogorov-Arnold Networks (KANs) can perform quantum state tomography while remaining interpretable, recovering physical structure without superior performance. The method identifies relevant Pauli measurements from 63 total measurements and reveals internal pathways consistent with known quantum mechanics, validating that neural models can be audited against established physics.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers introduce Inference-Time Argumentation (ITA), a neurosymbolic framework that combines large language models with formal argumentation semantics for claim verification. The system generates arguments, scores them, and produces ternary (true/false/uncertain) predictions with faithful, inspectable reasoning structures rather than post-hoc justifications.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers introduce ExtraCare, a domain adaptation method for clinical AI models that decomposes patient data into interpretable components while maintaining prediction accuracy across different healthcare datasets. The approach addresses a critical gap in healthcare AI by combining superior performance with transparent, explainable outputs—essential for clinical adoption where transparency and safety are paramount.
AIBullisharXiv – CS AI · Jun 106/10
🧠Researchers propose SAFE, an LLM-as-verifier framework that improves multi-hop question answering by validating reasoning steps against evidence during generation rather than only checking final answers. The approach uses Knowledge Graph triples to decompose reasoning into verifiable units and achieves 8.8 percentage point accuracy improvements across three benchmarks.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers introduce Regimes, an auditable autonomous improvement loop built on the ActiveGraph event-sourced runtime that enables transparent, reproducible AI agent optimization. The system diagnoses failures, proposes repairs, and validates them through multiple gates before promotion, demonstrating 5-10% held-out accuracy improvements on long-context reading comprehension tasks.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers introduce DiRL, a reinforcement learning framework that distinguishes between genuine reasoning and memorization in large language models by anchoring exploration to an internal reasoning-memorization direction. The method integrates with Group Relative Policy Optimization to improve performance on mathematical and reasoning benchmarks while suppressing exploration of memorized shortcuts.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers find that large language models make decisions based on systematic behavioral patterns but struggle to accurately articulate their reasoning. The study reveals a disconnect between what LLMs claim influences their choices and the attributes that actually drive their decisions, suggesting models operate with 'superficial beliefs' rather than fully understood decision frameworks.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers introduce WorldModelLens, an open-source interpretability framework that unifies analysis across diverse world model architectures (recurrent state-space models, token-based transformers, and joint-embedding systems) through a standardized capability-typed interface. The tool enables researchers to apply interpretability methods once rather than reimplementing them for each model architecture, addressing fragmentation in AI model analysis tooling.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers introduce a framework using compact proofs to measure feature interactions in crosscoders and Sparse Autoencoders, revealing that interactions between learned features cause reconstruction errors. The work demonstrates practical applications including computationally sparse models that maintain 60% performance with minimal features and detection of sleeper agent behavior in AI systems.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers have developed sparse autoencoders to interpret and control how language models process text-to-speech synthesis in CosyVoice3. The work demonstrates that interpretable features—phonemes, laughter, accent, and speaker gender—are causally linked to speech output and can be precisely steered to modify synthesis behavior without retraining.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers challenge the conventional wisdom that information leakage in concept-based neural networks is inherently harmful, arguing that some leakage is necessary for building accurate and practical AI systems. The paper proposes that 'benign leakage' can coexist with interpretability when concept descriptions are incomplete, reframing how these models should be optimized.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers demonstrate that linear structures in neural networks exist locally rather than globally, with task-specific directions that evolve during training rather than remaining stationary. Their findings on transformer models and LoRA adapters suggest that parameter adjustment techniques like task vectors work through dynamic geometric patterns that partially align across weight and activation spaces.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers developed a self-evolving scientific agent powered by large language models that autonomously discovers interpretable control policies for complex physical systems. The system successfully solved an underactuated fluid-dynamics problem (dogfish swimmer navigation) by iteratively testing strategies, diagnosing behaviors, and refining source code—achieving generalization to unseen targets without retraining.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers propose a variability-based framework for automatically naming concepts generated by Formal Concept Analysis (FCA) and Relational Concept Analysis (RCA) using large language models. The framework addresses the challenge of translating formally-defined but opaque symbolic abstractions into human-readable names by controlling which information sources (intent, extent, implications, relations) are exposed during naming, making semantic choices explicit and interpretable.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers propose a novel method for explaining black-box language model predictions by identifying linguistically-structured word subsets without requiring access to internal model parameters or gradients. The approach uses reinforcement learning and graph-based linguistic knowledge to generate interpretable, efficient explanations that outperform existing methods across multiple architectures and datasets.
AINeutralarXiv – CS AI · Jun 96/10
🧠REFLECT is a new method for identifying errors in long reasoning traces produced by LLM agents, particularly addressing the challenging "silent failure" problem where outputs appear plausible but are incorrect. The approach improves upon existing error-localization techniques by using controlled replay and contrastive evidence to refine error attribution, achieving higher accuracy across multiple benchmarks without requiring ground-truth answers.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers introduce ABLE, a framework that represents and compares large language models through gradient-based feature attributions rather than parameter analysis or output comparison. The training-free method achieves competitive performance on model comparison tasks across 239 open-source LLMs while providing theoretical stability guarantees.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers introduce mllm-shap, an open-source framework that extends Shapley Value explainability techniques to multimodal large language models processing text and audio inputs simultaneously. The platform addresses three technical challenges unique to multimodal systems and implements five estimation strategies, with a novel phonetic alignment technique reducing computational complexity by 10-50x.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers have developed a novel framework extending Shapley Values—a traditional explainability method—to multimodal large language models that process both text and audio. The work introduces computational optimizations and a preprocessing technique called Spectrogram-Guided Phonetic Alignment to make the analysis feasible, alongside an open-source tool for visualization, revealing that input modality significantly affects model attribution patterns.
AINeutralarXiv – CS AI · Jun 96/10
🧠Query Lens extends the Logit Lens technique to improve the interpretability of sparse autoencoders by analyzing both encoder key features and decoder value features, while accounting for indirect downstream effects. The research introduces the Subspace Channel Hypothesis, suggesting that neural modules process features through layer-specific subspaces, advancing understanding of how AI models process and manipulate information.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers have developed a method to detect emergent misalignment in large language models during finetuning by monitoring internal representational shifts rather than relying solely on behavioral evaluation. The technique identifies dangerous model behavior through a low-dimensional geometric signature in activation space, achieving high detection accuracy with minimal computational overhead.
AINeutralarXiv – CS AI · Jun 96/10
🧠NeuroAlign presents a hierarchical machine learning framework that fuses functional MRI and diffusion tensor imaging data to improve detection of mild cognitive impairment. The system introduces novel alignment and interaction mechanisms between multimodal neuroimaging datasets, with a new attribution method for interpretability, demonstrating competitive results across multiple medical imaging datasets.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers present Causal Agent Replay (CAR), a new method for diagnosing why large language model agents fail by identifying which decision step caused a failure rather than just which action executed it. Using structural causal models and intervention-based analysis, CAR achieves significantly higher attribution accuracy than existing LLM-judge approaches and provides confidence-bounded explanations for agent failures.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers introduce TRIAGE, an LLM-based framework that uses dialectical reasoning to improve risk prediction on irregularly sampled medical time series data. The approach generates competing clinical outcome rationales to produce calibrated, continuous risk scores rather than overconfident binary predictions, achieving 3.3% AUPRC improvement and 81% reduction in calibration error versus baseline methods.
