Large Language Models (LLMs) and more broadly Artificial Intelligence (AI) systems are often described and understood in human-like terms, a phenomenon known as \emph{anthropomorphism}. This paper provides a synthesis of recent literature on anthropomorphism in AI, covering theoretical frameworks, the role of language...
Ismael T. Freire, Marceau Nahon, Maud van Lier et al.· 0 citations
Post-hoc context erasing over the KV cache is challenging because a local edit has a global consequence: once a span has been processed, its influence propagates into the cached states of all subsequent tokens. This issue arises naturally in long-context LLM applications, where stale, incorrect, or harmful context may...
Mufei Li, Shikun Liu, Dongqi Fu et al.· 0 citations
Weak-to-strong (W2S) generalization is a promising framework for scalable oversight, yet existing evaluations often test students under matched train-test distributions. Therefore, we study W2S preference learning under zero-shot distribution shift and find that strong students trained on weak preference labels can app...
Khoi Le, Tri Cao, Phong Nguyen et al.· 0 citations
It is increasingly important to evaluate the characteristics of systems based on large language models (LLMs). Evaluations in this context often rely on a curated benchmark set of input prompts provided to the LLM, where the output for each prompt may be assigned a binary or ordinal score and the aggregation of scores...
Saatvik Kher, Shang Wu, Rachel Longjohn et al.· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Large language models (LLMs) routinely fail to output the correct option in multiple-choice question answering (MCQA) while encoding the answer internally. We expose this latent knowledge via the Query--Key (QK) score, defined for an attention head as the inner product between the last-token query and the key at the en...
Eduard Tulchinskii, Kristian Kuznetsov, Laida Kushnareva et al.· 0 citations
Parallel thinking has enjoyed great success for boosting LLM performance on reasoning tasks without the need for any re-training. However, existing methods follow a think-first-then-decide paradigm, i.e., they first sample multiple reasoning paths, which inevitably leads to overgeneration, then prune or stop unnecessar...
Zena Al-Khalili, Rafi Hakim, Dietrich Klakow et al.· 0 citations
Data-centric training has emerged as a promising direction for improving large language models (LLMs) by optimizing not only model parameters but also the selection, composition, and weighting of training data during optimization. However, existing approaches to data selection, data mixture optimization, and data rewei...
Hao Liang, Zhengyang Zhao, Mingrui Chen et al.· 0 citations
A new scaling law with separate benefit and harm terms that allows the value of an AI token to change sign while also reducing to Chinchilla in the absence of AI text is proposed.
Jenna Russell, Ben Glickenhaus, Katherine Thai et al.· 0 citations
Converting free-text radiology reports into structured labels supports cohort building, quality assurance, and monitoring of clinical imaging models, but the strongest label extractors are hosted proprietary models whose use raises privacy, cost, and reproducibility concerns. We asked whether a fine-tuned open-weight m...
Aawez Mansuri, Kush Mehta, Mohammadreza Chavoshi et al.· 0 citations
The policy-gradient theorem gives the exact gradient under the current policy, but finite on-policy samples may miss rare high-return trajectories. We study whether tree search improves their coverage within a fixed budget while controlling gradient bias. We introduce On-Policy Parallel Tree Search (OPTS) and Tree Traj...
Jun-Yu Lu, Shi-Chao Weng, Zhi-Qiang Wang et al.· 0 citations
Pre-training equips large language models (LLMs) with a broad repertoire of behavioral patterns associated with roles, styles, values, and goals. Post-training teaches conditional enactment and makes a helpful Assistant the default, but it does not erase alternative modes from the weights; explicit prompts can therefor...
Ke-Mou Li, Zhuan Shi, Qi-Zhou Wang et al.· 0 citations
On-policy self-distillation (OPSD) trains mathematical reasoning models using a privileged teacher that sees a reference solution and supervises student-sampled prefixes. Standard OPSD uses one fixed parameter setting at every state, but nearby settings may offer additional supervision. We find that local parameter per...
Xin Wei, Yi-Fan Ding, Yoshua Y. Li et al.· 0 citations
A new method, called CW-Net, translates the reasoning process of an autonomous vehicle’s AI system into understandable concepts that explain its behavior.
MIT News · Artificial Intelligence· news.mit.eduAug 31, 2026
With millions of users across the world, Julia has been used to conduct cutting-edge research and to design new drugs, jet engines, heat pumps, and more.
