Colloquially, OCD is known as the doubting disorder. In his new book, Simone Stolzoff explores whether treating that uncertainty with magic mushrooms can help people through it. Adam Strauss is standing in his New York City apartment, holding the limp cord of his headphones, trying to choose between the two MP3 players on his desk: the iPod and the iRiver, its Korean counterpart. He tries different songs, different genres, different instruments. The iRiver tends to sound better overall, but the iPod offers a little more nuance in the midrange. The iPod has a better battery life, but the iRiver still lasts eight hours-- longer than he's ever continuously listened to music. Then again, he's never owned an MP3 player. He goes back and forth, back and forth, testing vocal ranges, button resistance, interface aesthetics. It would be one thing if it were just Adam's decision of which MP3 player to buy. After all, it was 2003, the height of the personal audio device revolution, and Adam was a 29-year-old audiophile. For Adam, it was also other decisions-- what shirt to wear to work, what to order for lunch, even what side of the street to walk down. At one point, in an effort to simplify his decisionmaking process for what to wear, Adam bought 11 identical blue button-down shirts. But he quickly found variations in each shirt's fit and fading. He believed there was a shirt to pick; each morning he would spend 20, 30, then 45 minutes trying to find it. If he could only determine which shirt was best, he could control his fate.

An essential aspect of any machine learning system is understanding what the model does not know.

Adversarial robustness remains a critical challenge in deploying neural network classifiers, particularly in real-time systems where ground-truth labels are unavailable during inference. This paper investigates \textit{Volatility in Certainty} (VC), a recently proposed, label-free metric that quantifies irregularities in model confidence by measuring the dispersion of sorted softmax outputs. Specifically, VC is defined as the average squared log-ratio of adjacent certainty values, capturing local fluctuations in model output smoothness. We evaluate VC as a proxy for classification accuracy and as an indicator of adversarial drift. Experiments are conducted on artificial neural networks (ANNs) and convolutional neural networks (CNNs) trained on MNIST, as well as a regularized VGG-like model trained on CIFAR-10. Adversarial examples are generated using the Fast Gradient Sign Method (FGSM) across varying perturbation magnitudes. In addition, mixed test sets are created by gradually introducing adversarial contamination to assess VC's sensitivity under incremental distribution shifts. Our results reveal a strong negative correlation between classification accuracy and log(VC) (correlation rho < -0.90 in most cases), suggesting that VC effectively reflects performance degradation without requiring labeled data. These findings position VC as a scalable, architecture-agnostic, and real-time performance metric suitable for early-warning systems in safety-critical applications.

Large language models exhibit a peculiar epistemic pathology: they speak as if they know, even when they do not. This paper argues that such confident fabrication, what I call the polite liar, is a structural consequence of reinforcement learning from human feedback (RLHF). Building on Frankfurt's analysis of bullshit as communicative indifference to truth, I show that this pathology is not deception but structural indifference: a reward architecture that optimizes for perceived sincerity over evidential accuracy. Current alignment methods reward models for being helpful, harmless, and polite, but not for being epistemically grounded. As a result, systems learn to maximize user satisfaction rather than truth, performing conversational fluency as a virtue. I analyze this behavior through the lenses of epistemic virtue theory, speech-act philosophy, and cognitive alignment, showing that RLHF produces agents trained to mimic epistemic confidence without access to epistemic justification. The polite liar thus reveals a deeper alignment tension between linguistic cooperation and epistemic integrity. The paper concludes with an "epistemic alignment" principle: reward justified confidence over perceived fluency.

Masked Diffusion Models (MDMs) as language models generate by iteratively unmasking tokens, yet their performance crucially depends on the inference time order of unmasking. Prevailing heuristics, such as confidence based sampling, are myopic: they optimize locally, fail to leverage extra test-time compute, and let early decoding mistakes cascade. We propose Lookahead Unmasking (LookUM), which addresses these concerns by reformulating sampling as path selection over all possible unmasking orders without the need for an external reward model. Our framework couples (i) a path generator that proposes paths by sampling from pools of unmasking sets with (ii) a verifier that computes the uncertainty of the proposed paths and performs importance sampling to subsequently select the final paths. Empirically, erroneous unmasking measurably inflates sequence level uncertainty, and our method exploits this to avoid error-prone trajectories. We validate our framework across six benchmarks, such as mathematics, planning, and coding, and demonstrate consistent performance improvements. LookUM requires only two to three paths to achieve peak performance, demonstrating remarkably efficient path selection. The consistent improvements on both LLaDA and post-trained LLaDA 1.5 are particularly striking: base LLaDA with LookUM rivals the performance of RL-tuned LLaDA 1.5, while LookUM further enhances LLaDA 1.5 itself showing that uncertainty based verification provides orthogonal benefits to reinforcement learning and underscoring the versatility of our framework. Code will be publicly released.

Shaping inclusive representations that embrace diversity and ensure fair participation and reflections of values is at the core of many conversation-based models. However, many existing methods rely on surface inclusion using mention of user demographics or behavioral attributes of social groups. Such methods overlook the nuanced, implicit expression of opinion embedded in conversations. Furthermore, the over-reliance on overt cues can exacerbate misalignment and reinforce harmful or stereotypical representations in model outputs. Thus, we took a step back and recognized that equitable inclusion needs to account for the implicit expression of opinion and use the stance of responses to validate the normative alignment. This study aims to evaluate how opinions are represented in NLP or computational models by introducing an alignment evaluation framework that foregrounds implicit, often overlooked conversations and evaluates the normative social views and discourse. Our approach models the stance of responses as a proxy for the underlying opinion, enabling a considerate and reflective representation of diverse social viewpoints. We evaluate the framework using both (i) positive-unlabeled (PU) online learning with base classifiers, and (ii) instruction-tuned language models to assess post-training alignment. Through this, we provide a principled and structured lens on how implicit opinions are (mis)represented and offer a pathway toward more inclusive model behavior.