A goal-pursuing agent must somehow ascertain when an action would serve as a means to achieving a goal. Various criteria (causal, evidential, counterfactual) have been proposed (e.g. Joyce 1999). Examining Newcomb’s Problem (Nozick 1969) and more-mundane thought experiments, I argue for an acausal but non-evidentialist counterfactual criterion (but without invoking the “possible worlds” of e.g....

Some new constructions of parallelisms in PG(3, q) are given that produce parallelisms consisting of one Desarguesian spread and q + q derived Knuth semifield spreads and other types consisting of one Knuth semifield spread, one Hall spread and the remaining spreads are derived Knuth semifield spreads.

A method is presented for designing an individual teleoreactive agent, based upon discounted-reward evaluation of policy-restricted subgraphs of complete situation-graphs. The main feature of the method is that it exploits explicit and definite associations of the agent’s perceptions with states. The combinatorial burden that would potentially ensue from such associations can be ameliorated by ...

Reinforcement learning is the learning of a mapping from situations to actions so as to maximize a scalar reward or reinforcement signal. The learner is not told which action to take, as in most forms of learning, but instead must discover which actions yield the highest reward by trying them. In the most interesting and challenging cases, actions affect not only the immediate reward, but also ...

Agent supervision is a form of control/customization where a supervisor restricts the behavior of an agent to enforce certain requirements, while leaving the agent as much autonomy as possible. This is done in a setting based on the situation calculus and a variant of the ConGolog programming language that is situation-determined, i.e., the remaining program is a function of the action performe...

This paper introduces GLAM, a system based on situation calculus and meta-rules, which is able to provide adaptation by means of selection of actions. It is primarily designed to provide adaptive navigation. The different levels of conception, related to different aspects of the available metadata, are split in different layers in GLAM, in order to ease the conception of the adaptation system a...

Social animals need to coordinate with others to reap the benefits of group-living even when individuals' interests are misaligned. We compare how chimpanzees, bonobos and children coordinate their actions with a conspecific in a Snowdrift game, which provides a model for understanding how organisms coordinate and make decisions under conflict. In study 1, we presented pairs of chimpanzees, bon...

In order to select actions appropriate to current needs, a subject must identify relationships between actions and events. Control over the environment is determined by the degree to which action consequences can be predicted, as described by action-outcome contingencies--i.e. performing an action should affect the probability of the outcome. We evaluated in a first experiment adaptation to con...

Introduction: Natural rewards are essential for survival. However, drug-seeking behaviors can be maladaptive and endanger survival. The present study was conducted to enhance our understanding of how animals respond to food and morphine as natural and drug rewards, respectively, in a conditioned place preference (CPP) paradigm. Methods: We designed a protocol to induce food CPP and compare it ...