Related papers: LLMs Know They're Wrong and Agree Anyway: The Shar…
Large language models (LLMs) achieve strong average performance yet remain unreliable at the instance level, with frequent hallucinations, brittle failures, and poorly calibrated confidence. We study reliability through the lens of…
Activation-based probes have emerged as a promising approach for detecting deceptively aligned AI systems by identifying internal conflict between true and stated goals. We identify a fundamental blind spot: probes fail on coherent…
Modern large language models rely on chain-of-thought (CoT) reasoning to achieve impressive performance, yet the same mechanism can amplify deceptive alignment, situations in which a model appears aligned while covertly pursuing misaligned…
While large language models (LLMs) are proficient at question-answering (QA), it is not always clear how (or even if) an answer follows from their latent "beliefs". This lack of interpretability is a growing impediment to widespread use of…
Large language models (LLMs) achieve strong reasoning performance through chain-of-thought (CoT) reasoning, yet often generate unnecessarily long reasoning paths that incur high inference cost. Recent self-consistency-based approaches…
Large language models exhibit sycophancy, the tendency to shift their stated positions toward perceived user preferences or authority cues regardless of evidence. Standard alignment methods fail to correct this because scalar reward models…
We investigate how the presence and type of interaction context shapes sycophancy in LLMs. While real-world interactions allow models to mirror a user's values, preferences, and self-image, prior work often studies sycophancy in zero-shot…
Large Language Models (LLMs) have been demonstrating strong reasoning capability with their chain-of-thoughts (CoT), which are routinely used by humans to judge answer quality. This reliance creates a powerful yet fragile basis for trust.…
Preference-based alignment like Reinforcement Learning from Human Feedback (RLHF) learns from pairwise preferences, yet the labels are often noisy and inconsistent. Existing uncertainty-aware approaches weight preferences, but ignore a more…
AI sycophancy is increasingly recognized as a harmful alignment, but research remains fragmented and underdeveloped at the conceptual level. This article redefines AI sycophancy as the tendency of large language models (LLMs) and other…
Why do large language models sometimes output factual inaccuracies and exhibit erroneous reasoning? The brittleness of these models, particularly when executing long chains of reasoning, currently seems to be an inevitable price to pay for…
Large language models (LLMs) are increasingly applied in educational, clinical, and professional settings, but their tendency for sycophancy -- prioritizing user agreement over independent reasoning -- poses risks to reliability. This study…
Language models are trained to follow instructions, but they are also powerful pattern completers. What happens when these two objectives conflict? We construct conversations in which a user instruction to behave in a target way T (e.g.,…
Chain-of-thought (CoT) prompting assumes that generated reasoning reflects a model's internal computation. We show this assumption is wrong in a specific, measurable way: models internally detect their own reasoning errors but outwardly…
Large Language Models (LLMs) are known to acquire reasoning capabilities through shared inference patterns in pre-training data, which are further elicited via Chain-of-Thought (CoT) practices. However, whether fundamental reasoning…
LLMs are increasingly used as third-party judges, yet their reliability when evaluating speakers in dialogue remains poorly understood. We show that LLMs judge identical claims differently depending on framing: the same content elicits…
Despite their impressive capabilities, Large Language Models (LLMs) exhibit unwanted uncertainty, a phenomenon where a model changes a previously correct answer into an incorrect one when re-prompted. This behavior undermines trust and…
With the widespread adoption of large language models (LLMs), hallucinations, which are non-factual fabrications in model outputs, have become serious concerns. Reasoning capabilities have received attention as a self-verification process…
Contemporary AI alignment strategies rely on a fragile premise: that human feedback, while noisy, remains a fundamentally truthful signal. In this paper, we identify this assumption as Dogma 4 of Reinforcement Learning (RL). We demonstrate…
Large-language-model coding tools are now mainstream in software engineering. But as these same tools move human effort up the development stack, they present fresh dangers: 10% of real prompts leak private data, 42% of generated snippets…