Three Roles, One Model: Role Orchestration at Inference Time to Close the Performance Gap Between Small and Large Agents

Large language model (LLM) agents show promise on realistic tool-use tasks, but deploying capable agents on modest hardware remains challenging. We study whether inference-time scaffolding alone, without any additional training compute, can improve the performance of a small model in complex multi-step environments. Operating on a single 24\,GB GPU, we evaluate Qwen3-8B under both full-precision (FP16, 12K context) and 4-bit quantized (AWQ, 32K context) configurations. Without any intervention, the raw model achieves just 5.4\% (FP16) and 3.0\% (AWQ) task goal completion. Guided by a systematic failure mode analysis, we introduce a three-tier inference scaffolding pipeline that deploys the same frozen model in three distinct roles: (1) a summarization model that preserves critical artifacts (tokens, credentials, API responses) while compressing dialogue history; (2) the main agent model that reasons over the compressed context; and (3) an isolated correction model that reviews and revises the agent's code output without access to conversation history, breaking repetitive failure loops. Applied to the same unmodified model, this scaffolding yields 8.9\% (FP16) and 5.9\% (AWQ) task goal completion, roughly doubling performance in both settings, with particularly strong gains on difficulty-1 tasks (15.8\%$\to$26.3\% FP16; 5.3\%$\to$14.0\% AWQ). On full-precision inference, our scaffolded 8B model surpasses DeepSeek-Coder 33B Instruct (7.1\%) from the original AppWorld evaluation, demonstrating that structured inference-time interventions can make small models competitive with systems 4$\times$ their size. We formalize the approach as a scaffolded policy over a frozen base model, three invocations of the same weights with different conditioning, drawing connections to test-time compute scaling and action-space shaping in reinforcement learning.

Gauging Public Acceptance of Conditionally Automated Cars in the United States

In this work we look at an element of smart cities, conditionally automated cars (SAE Level 3), investigating the factors influencing public acceptance in the United States. We apply an adaptation of the UTUAT2 model. Taking an experimental approach study 358 participants in the US were presented with a vignette outlining the L3 technology followed by a series of questions to capture their perceptions of conditionally automated cars. PLS-SEM was used to analyze the collected data. The results reveal that the acceptance of the technology, in order of decreasing importance, was determined by social influence, performance expectancy, hedonic motivation, facilitating conditions, and effort expectancy. Furthermore, hedonic motivation, social influence, facilitating conditions and effort expectancy all have a positive influence on the perception of how useful the technology is; facilitating conditions, hedonic motivation, and social influence all have a positive influence on effort expectancy; social influence and facilitating conditions positively influence hedonic motivation; and social influence positively influences facilitating conditions. A moderating effect for gender was found, with the effect of hedonic motivation influencing intention to adopt is more prominent for men.

Investigating End-user Acceptance of Last-mile Delivery by Autonomous Vehicles in the United States

This paper investigates the end-user acceptance of last-mile delivery carried out by autonomous vehicles within the United States. A total of 296 participants were presented with information on this technology and then asked to complete a questionnaire on their perceptions to gauge their behavioral intention concerning acceptance. Structural equation modeling of the partial least squares flavor (PLS-SEM) was employed to analyze the collected data. The results indicated that the perceived usefulness of the technology played the greatest role in end-user acceptance decisions, followed by the influence of others, and then the enjoyment received by interacting with the technology. Furthermore, the perception of risk associated with using autonomous delivery vehicles for last-mile delivery led to a decrease in acceptance. However, most participants did not perceive the use of this technology to be risky. The paper concludes by summarizing the implications our findings have on the respective stakeholders and proposing the next steps in this area of research.