Model releases, architecture work and prompting research on large language models — from frontier-lab announcements to the arXiv papers behind them.
arXiv:2603. 09995v2 Announce Type: replace-cross Abstract: Behavioral interview evaluation using large language models presents unique challenges that require structured assessment, realistic interviewer behavior simulation, and pedagogical value for candidate training.
arXiv:2606. 08768v1 Announce Type: new Abstract: Transformers consistently fail to learn certain simple functions that are provably expressible with specific parameter settings.
arXiv:2606. 09456v1 Announce Type: new Abstract: On-Policy Distillation (OPD) has become a core technique in the post-training of Large Language Models (LLMs) for transferring knowledge from domain experts to student models.
arXiv:2605. 28207v2 Announce Type: replace-cross Abstract: Mixture-of-Experts (MoE) is now the dominant architecture for frontier language models, yet it requires all expert parameters to be loaded in memory, making it less preferable for memory-constrained deployment.
arXiv:2402. 08922v3 Announce Type: replace Abstract: Large-scale black-box models have become ubiquitous across numerous applications.
arXiv:2605. 24942v2 Announce Type: replace-cross Abstract: Steering a language model - intervening on its internal activations to change downstream behaviour - has recently expanded beyond linear interpolation to nonlinear methods such as angular and kernelized steering, which define intervention transformations without learning an explicit geometry over paths in activation space.
arXiv:2606. 08169v1 Announce Type: cross Abstract: Enabling robots to understand and execute tasks from natural language commands while maintaining data efficiency remains challenging.
arXiv:2606. 07521v1 Announce Type: cross Abstract: This study investigates the phenomenon of hallucinations in domain-adapted Large Language Models (LLMs), focusing on the fine-tuning of the Llama-2 model with the Lamini dataset.
arXiv:2605. 00327v2 Announce Type: replace-cross Abstract: In large language model (LLM)-based recommendation systems, direct preference optimization (DPO) effectively aligns recommendations with user preferences, requiring multi-negative objective functions to leverage abundant implicit-feedback negatives and sharpen preference boundaries.
arXiv:2606. 07690v1 Announce Type: cross Abstract: Finetuning data selection requires balancing two competing goals: selecting examples that improve the downstream objective, and doing so without repeatedly finetuning models.
arXiv:2606. 08892v1 Announce Type: new Abstract: AI models deployed in critical domains, such as AI safety research, may subtly sabotage our efforts due to misalignment.
arXiv:2606. 07857v1 Announce Type: cross Abstract: The rise of edge-based machine learning has enabled distributed adaptation of language models across mobile and IoT devices, offering privacy preservation and real-time responsiveness.
arXiv:2606. 09500v1 Announce Type: new Abstract: Objective.
arXiv:2606. 08779v1 Announce Type: new Abstract: Reinforcement Learning (RL) has emerged as a pivotal post-training paradigm, yet it frequently suffers from unpredictable sub-optimum performance or even training collapses.
arXiv:2509. 22097v5 Announce Type: replace-cross Abstract: Large language model-powered code agents are rapidly transforming software engineering, yet the security risks of their generated code have become a critical concern.
arXiv:2510. 27544v2 Announce Type: replace Abstract: Temporal reasoning involves understanding how systems evolve over time through input-driven state transitions.
arXiv:2310. 10196v3 Announce Type: replace-cross Abstract: Temporal data, including time series and spatio-temporal data, are pervasive in real-world applications.
arXiv:2606. 09157v1 Announce Type: cross Abstract: This paper revisits our pipeline called Syllogistic Evaluation Framework-Common Logic Grammar Construction (SEF-CLGC).
arXiv:2602. 02572v2 Announce Type: replace-cross Abstract: Existing alignment methods directly use the reward model learned from user preference data to optimize an LLM policy, subject to KL regularization with respect to the base policy.
arXiv:2605. 11484v2 Announce Type: replace Abstract: Task completion in digital and physical environments increasingly involves complex temporal interaction, where actions and observations unfold over different time scales rather than align with fixed observation--action steps.