Scorio.jl is a Julia package for evaluating and ranking systems from repeated responses to shared tasks. It provides a common tensor-based interface for direct score-based, pairwise, psychometric, voting, graph, and listwise methods, so the same benchmark can be analyzed under multiple ranking assumptions. We describe the package design, position it relative to existing Julia tools, and report pilot experiments on synthetic rank recovery, stability under limited trials, and runtime scaling.

We further design an expert-wise training strategy to alleviate the impact of unhelpful demonstrations when optimizing the retriever model.

Partial observability of the underlying states generally presents significant challenges for reinforcement learning (RL).

Asabaseline for DCLM, we conduct extensive experiments and find that model-based filtering is key to assembling a high-quality training set.

War of 1812 veteran Shadrack Byfield's second book describes a grittier life story--and a hook for a hand. Breakthroughs, discoveries, and DIY tips sent six days a week. A long-lost second memoir penned by a famed 19th-century British soldier named Shadlock Byfield resurfaced in a rather unexpected place--Cleveland, Ohio. As explained in a study recently published in the, Byfield's second book depicts a very different war veteran than the one described in his first autobiography written 11 years earlier. Although he may not be a household name, many early American history buffs are well acquainted with Shadrack Byfield .

Large language models (LLMs) excel at complex reasoning tasks but remain computationally expensive, limiting their practical deployment. To address this, recent works have focused on distilling reasoning capabilities into smaller language models (sLMs) using chain-of-thought (CoT) traces from teacher LLMs. However, this approach struggles in scenarios requiring rare factual knowledge or precise computation, where sLMs often hallucinate due to limited capability. In this work, we propose Agent Distillation, a framework for transferring not only reasoning capability but full task-solving behavior from LLM-based agents into sLMs with retrieval and code tools. We improve agent distillation along two complementary axes: (1) we introduce a prompting method called first-thought prefix to enhance the quality of teacher-generated trajectories; and (2) we propose a self-consistent action generation for improving test-time robustness of small agents. We evaluate our method on eight reasoning tasks across factual and mathematical domains, covering both in-domain and out-of-domain generalization. Our results show that sLMs as small as 0.5B, 1.5B, 3B parameters can achieve performance competitive with next-tier larger 1.5B, 3B, 7B models fine-tuned using CoT distillation, demonstrating the potential of agent distillation for building practical, tool-using small agents. Our code is available at https://github.com/Nardien/agent-distillation.