Finetuning allows us to adapt pretrained LLMs in a cost-efficient manner. But which method should we use? This article compares different parameter-efficient finetuning methods for the latest top-performing open-source LLM, Falcon. Using parameter-efficient finetuning methods outlined in this article, it’s possible to finetune an LLM in 1 hour on a single GPU instead of a day on 6 GPUs.

Ver mais

Reinforcement learning is the method of choice to train models in sampling-based setups with binary outcome feedback, such as navigation, code generation, and mathematical problem solving. In such settings, models implicitly induce a likelihood over correct rollouts. However, we observe that reinforcement learning does not maximize this likelihood, and instead optimizes only a lower-order approximation. Inspired by this observation, we introduce Maximum Likelihood Reinforcement Learning (MaxRL), a sampling-based framework to approximate maximum likelihood using reinforcement learning techniques. MaxRL […]

Ver mais

arXiv:2602.04985v1 Announce Type: new Abstract: Timely delivery and optimal routing remain fundamental challenges in the modern logistics industry. Building on prior work that considers single-package delivery across networks using multiple types of collaborative agents with restricted movement areas (e.g., drones or trucks), we examine the complexity of the problem under structural and operational constraints. Our focus is on minimizing total delivery time by coordinating agents that differ in speed and movement range across a graph. This problem formulation […]

Ver mais

Force and torque (F/T) sensing is critical for robot-environment interaction, but physical F/T sensors impose constraints in size, cost, and fragility. To mitigate this, recent studies have estimated force/wrench sensorlessly from robot internal states. While existing methods generally target relatively slow interactions, tasks involving rapid interactions, such as grinding, can induce task-critical high-frequency vibrations, and estimation in such robotic settings remains underexplored. To address this gap, we propose a Frequency-aware Decomposition Network (FDN) for short-term forecasting of vibration-rich […]

Ver mais

arXiv:2603.23518v1 Announce Type: new Abstract: General-purpose embedding models excel at recognizing semantic similarities but fail to capture the characteristics of texts specified by user instructions. In contrast, instruction-tuned embedders can align embeddings with textual instructions yet cannot autonomously infer latent corpus structures, such as determining the optimal number of clusters. To address both limitations, we reframe instruction-following clustering as a generative task and train large reasoning models (LRMs) as autonomous clustering agents. Our reasoning-driven training pipeline enables LRMs […]

Ver mais

I’m building MH-FLOCKE, an open embodied AI framework that replaces standard RL with biologically grounded learning. The goal: a reusable platform where spiking neural networks, cerebellar models, and predictive coding work together — not as isolated papers but as one integrated system. The robot is a Unitree Go2 in MuJoCo, controlled by a spiking neural network (4,624 Izhikevich neurons, 93k synapses) with a Marr-Albus-Ito cerebellar forward model. The learning signal is not a shaped reward. Instead: Task-specific Prediction […]

Ver mais

arXiv:2603.12287v1 Announce Type: new Abstract: We address the problem of transforming raw vessel trajectory data collected from AIS into structured and semantically enriched representations interpretable by humans and directly usable by machine reasoning systems. We propose a context-aware trajectory abstraction framework that segments noisy AIS sequences into distinct trips each consisting of clean, mobility-annotated episodes. Each episode is further enriched with multi-source contextual information, such as nearby geographic entities, offshore navigation features, and weather conditions. Crucially, such representations […]

Ver mais

Policy steering is an emerging way to adapt robot behaviors at deployment-time: a learned verifier analyzes low-level action samples proposed by a pre-trained policy (e.g., diffusion policy) and selects only those aligned with the task. While Vision-Language Models (VLMs) are promising general-purpose verifiers due to their reasoning capabilities, existing frameworks often assume these models are well-calibrated. In practice, the overconfident judgment from VLM can degrade the steering performance under both high-level semantic uncertainty in task specifications and low-level […]

Ver mais