44 articles mentioning Lens
arXiv:2610.28187v1 Announce Type: cross Abstract: Testing LLM-based software is expensive: each input needs several sampled responses, each judged by a costly oracle. Further, every prompt or model change calls for re-testing. Pre-generation adequacy measures cut this cost by ranking inputs by failu
arXiv:2610.01633v2 Announce Type: replace Abstract: Explaining generalization and training dynamics in neural networks remains a challenge, and various approaches have been developed to study different aspects of these phenomena. In this paper, we introduce the idea of {\em magnitude potential} -- a
arXiv:2609.27355v2 Announce Type: replace-cross Abstract: Unlearning ensures LLM compliance by removing the influence of private or copyrighted training data. However, since LLM models typically undergo post-training compression, like quantization, in practical deployment, it has been observed that
arXiv:2610.15900v1 Announce Type: cross Abstract: We present Gen Through the Lens, a camera-controllable image-to-video generation framework that synthesizes videos from a given reference image while enabling realistic and physically consistent camera operations. Unlike previous models that treat ca
arXiv:2511.17937v2 Announce Type: replace Abstract: Alignment faking is a form of strategic deception in AI in which models selectively comply with training objectives when they infer that they are in training, while preserving different behavior outside training. The phenomenon was first documented
arXiv:2506.23101v2 Announce Type: replace-cross Abstract: Ensuring the trustworthiness of Multimodal Large Language Models (MLLMs) requires rigorous evaluation of gender bias, particularly in socially sensitive domains. Existing benchmarks primarily assess bias in isolated contexts and are less suit
arXiv:2610.19827v1 Announce Type: cross Abstract: LLMs absorb vast knowledge from training data, yet their inherently probabilistic nature offers no guarantee that this knowledge will surface in their outputs. To bridge this gap, Knowledge Extraction Enhancement (KEE) techniques such as activation s
arXiv:2608.21389v2 Announce Type: replace-cross Abstract: Generative AI enables customized misinformation at scale, yet defenses remain largely reactive. We present empirical findings from a human-subject study (n=504 participants, n=2,438 judgments) in which users classified news fragments by origi
arXiv:2609.39263v1 Announce Type: new Abstract: A concept subspace's effect on model behavior does not establish how it relates to the output readout. We introduce a two-sided geometric diagnostic that measures an extracted subspace's overlap with the dominant right-singular directions of the unembe
arXiv:2609.38453v1 Announce Type: new Abstract: Grokking shows that fitting the training data and learning the underlying signal can occur at very different stages. However, existing theories offer limited quantitative insight into how this delayed generalization depends on inductive bias and signal
arXiv:2603.13587v2 Announce Type: replace-cross Abstract: State-space models (SSMs) are effective architectures for sequential modeling, but a rigorous theoretical understanding of their training dynamics is still lacking. We formulate continuous-time SSM parameter-path optimization as an ensemble o
arXiv:2610.00294v1 Announce Type: cross Abstract: Automated acne severity grading requires both whole-face context and fine-grained lesion evidence. We propose LENS-GRF (Lesion Evidence Network with Set-Transformer and Gated Residual Fusion), an interpretable multi-stage framework for four-class acn
arXiv:2609.39449v1 Announce Type: cross Abstract: Distributionally robust optimization (DRO) studies parameter estimation under uncertainty in the underlying probability distribution and has emerged as a principled framework for analyzing robustness and generalization. In particular, Wasserstein DRO
arXiv:2609.36262v1 Announce Type: new Abstract: Recent studies have observed that parameter changes during language-model post-training can be concentrated in a small subset of coordinates. This phenomenon has been reported in reinforcement learning, on-policy distillation, and supervised fine-tunin
arXiv:2602.24055v5 Announce Type: replace Abstract: This study proposes CIRCLE, a six-stage, lifecycle-based framework to bridge the reality gap between model-centric performance metrics and AI system outcomes in deployment. Current approaches such as MLOps frameworks and AI model benchmarks offer d
arXiv:2609.31360v1 Announce Type: new Abstract: Inference in LLMs is conventionally a fixed-depth, fixed-order forward pass through every layer, regardless of how difficult the input is. The human brain does not work this way: using the thalamus as a central hub, it routes information flexibly to al
arXiv:2602.12316v3 Announce Type: replace Abstract: Frontier AI systems are increasingly capable and deployed in high-stakes multi-agent environments. However, existing AI safety benchmarks largely evaluate single agents, leaving multi-agent risks such as coordination failure and conflict poorly und
arXiv:2609.20252v1 Announce Type: cross Abstract: High-quality representations are essential for a wide range of downstream tasks. Dedicated embedding models are explicitly optimized for representation learning, yet their training data are often more limited in scale and diversity than the massive c
arXiv:2609.05111v2 Announce Type: replace Abstract: Supervised fine-tuning (SFT), few-shot in-context learning (ICL), KL-regularized RLHF/RLVR, and on-policy distillation are usually treated as distinct post-training paradigms. We develop a unified Bayesian perspective in which each is an instance o
arXiv:2609.10084v1 Announce Type: cross Abstract: Generalized zero-shot learning (GZSL) has emerged as an important paradigm for visual recognition systems that must generalize to classes that were not observed during training. Traditional GZSL techniques are limited by their applicability to a rela
arXiv:2609.09124v1 Announce Type: cross Abstract: Visual encoders construct a representation of the image input for Vision-Language models. How much conceptual, as opposed to immediately visible, information does this representation contain? We use canonical color as a controlled test case to ask wh
arXiv:2609.05575v1 Announce Type: new Abstract: Understanding how concepts are encoded in the internal representations of machine learning models is a central problem in mechanistic interpretability, essential both for the science of deep learning and for the trustworthy deployment of increasingly c
arXiv:2609.01924v1 Announce Type: new Abstract: Recent work identifies a mid-depth band of verbalisable, causally potent representations in a standard feedforward transformer --- a functional analogue of a global workspace. Whether the same workspace functionality emerges when depth is implemented t
arXiv:2609.01936v1 Announce Type: cross Abstract: A language model's prediction of its next token develops across layers, and lens methods track this process by decoding intermediate hidden states into tokens. But a lens reading reflects both the hidden state and the readout (the unembedding matrix)
arXiv:2608.27910v2 Announce Type: replace Abstract: As large language models and increasingly capable AI agents are deployed in high-risk settings, aligning them with complex human values has become a central challenge. Existing alignment methods, while effective in improving helpfulness, harmlessne
arXiv:2608.31084v1 Announce Type: new Abstract: The Jacobian Lens (J-lens) is a recent tool for interpreting LLMs. It reads a hidden state as a ranked list of vocabulary tokens, leaving multi-token concepts without a representation of their own. The original J-lens work addresses this limitation wit
arXiv:2506.08311v3 Announce Type: replace-cross Abstract: Automated Program Repair (APR) agents leverage large language models (LLMs) to autonomously diagnose and patch software bugs using planning, reasoning, and tools. Although these agents show strong performance on leaderboards such as SWE-bench
arXiv:2608.27813v1 Announce Type: new Abstract: Structural probes were introduced by Hewitt and Manning to reconstruct syntactic trees from a neural language model's latent representations. They are evaluated by calculating the proportion of syntactic tree edges correctly reconstructed over an annot
arXiv:2608.27260v1 Announce Type: new Abstract: LLM agents increasingly rely on generated interaction data to learn how to interact with external environments. Agentic data generation must maintain consistency among environments, tasks, interactions, and success signals while producing experience th
arXiv:2603.21165v3 Announce Type: replace Abstract: Bangla culture is richly expressed through region, dialect, history, food, politics, media, and everyday visual life, yet it remains underrepresented in multimodal evaluation. To address this gap, we introduce BanglaVerse, a culturally grounded ben
arXiv:2608.25347v1 Announce Type: new Abstract: The Jacobian lens (J-lens) has been proposed as a way to read verbalizable representations from language models. However, its principle and meaning lack a detailed and theoretical discussion. We provide a mathematical view of this interpretation and of
arXiv:2607.22657v2 Announce Type: replace-cross Abstract: Large language models (LLMs) can reproduce disinformation-aligned narrative frames as plausible explanations, raising the question of whether existing machine-unlearning algorithms can suppress this behavior. We introduce Level-based Evaluati
arXiv:2410.06665v4 Announce Type: replace-cross Abstract: This paper explores the characterization of equivariant linear layers for representations of permutations and related groups. Unlike traditional approaches, which address these problems using parameter-sharing, we consider an alternative meth
arXiv:2607.23970v1 Announce Type: new Abstract: Machine Unlearning aims to remove undesired information from trained models without full retraining from scratch. Despite recent progress, the loss landscape and optimization geometry of unlearning are poorly understood. In this paper, we study machine
arXiv:2605.09159v2 Announce Type: replace Abstract: Recent work shows that large language models (LLMs) encode behavioral traits ("personas") as linear directions in activation space, often called "persona vectors". Prior work has used such directions as static handles for behavioral steering. We in
arXiv:2607.20241v1 Announce Type: cross Abstract: Culturally loaded translation poses unique challenges for machine translation (MT), as meanings are deeply embedded in socio-cultural contexts beyond surface linguistic forms. Although large language models (LLMs) have enabled MT systems to achieve h
arXiv:2606.08076v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) can generate high-quality arguments, yet their ability to engage in nuanced and persuasive communicative actions remains largely unexplored. This work explores the persuasive potential of LLMs through the framewor
arXiv:2605.31581v2 Announce Type: replace Abstract: The same arguments often need to be evaluated under different external regimes. An agent with influence over the regime has a strategic lever that standard formalisms do not directly capture. We introduce context-dependent argumentation frameworks
arXiv:2510.15824v2 Announce Type: replace-cross Abstract: This article considers an online version of conformal inference, called adaptive conformal inference [ACI] and introduced by Gibbs and Cand\`es (2021): prediction sets are issued sequentially, after observing features and before the outcomes
arXiv:2511.17126v5 Announce Type: replace-cross Abstract: Emerging deep-learning-based lens library pre-training (LensLib-PT) pipeline offers a new avenue for blind lens aberration correction by training a universal neural network, demonstrating strong capability in handling diverse unknown optical
arXiv:2607.07762v1 Announce Type: new Abstract: Modern machine learning (ML) increasingly relies on complex models whose behavior is difficult to characterize beyond empirical performance metrics. Across a wide range of tasks, including prediction, generation, and decision-making, models with simila
arXiv:2604.08580v2 Announce Type: replace-cross Abstract: Reward fine-tuning of diffusion and flow models and sampling from tilted or Boltzmann distributions can both be formulated as stochastic optimal control (SOC) problems, where learning an optimal generative dynamics corresponds to optimizing a
arXiv:2606.29693v2 Announce Type: replace Abstract: We ask a simple question about decoder-only transformers: between which two layers is the probability of a predicted token actually produced? Existing layer-wise readout tools answer only approximately. The logit lens and its trained variant report
arXiv:2606.30755v1 Announce Type: cross Abstract: Claw-like AI agents (e.g., OpenClaw) are always-on processes with persistent access to credentials, files, tools, and external services. They take on system-level responsibilities -- installing packages, maintaining state, scheduling subtasks, and me