State-of-the-art deep-learning systems use decision rules that are challenging for humans to model. Explainable AI (XAI) attempts to improve human understanding but rarely accounts for how people typically reason about unfamiliar agents. We propose explicitly modelling the human explainee via Bayesian teaching, which evaluates explanations by how much they shift explainees’ inferences toward a desired goal. We assess Bayesian teaching in a binary image classification task across a variety of contexts. Absent intervention, participants predict that the AI’s classifications will match their own, but explanations generated by Bayesian teaching improve their ability to predict the AI’s judgements by moving them away from this prior belief. Bayesian teaching further allows each case to be broken down into sub-examples (here saliency maps). These sub-examples complement whole examples by improving error detection for familiar categories, whereas whole examples help predict correct AI judgements of unfamiliar cases.

Probabilistic reinforcement learning declines in healthy cognitive aging. While some findings suggest impairments are especially conspicuous in learning from rewards, resembling deficits in Parkinson’s disease, others also show impairments in learning from punishments. To reconcile these findings, we tested 252 adults from three age groups on a probabilistic reinforcement learning task, analyzed trial-by-trial performance with a Q-reinforcement learning model, and correlated both fitted model parameters and behavior to polymorphisms in dopamine-related genes. Analyses revealed that learning from both positive and negative feedback declines with age, but through different mechanisms: When learning from negative feedback, older adults were slower due to noisy decision-making; when learning from positive feedback, they tended to settle for a non-optimal solution due to an imbalance in learning from positive and negative prediction errors. The imbalance was associated with polymorphisms in the DARPP-32 gene and appeared to arise from mechanisms different from those previously attributed to Parkinson’s disease. Moreover, this imbalance predicted previous findings on aging using the Probabilistic Selection Task, which were misattributed to Parkinsonian mechanisms.

In this study, investigation was carried out under in vitro as well as field conditions to explore inhibitors of sorghum grain mold. Phytochemicals, viz., methyl trans-p-coumarate (AIC-1), methyl caffeate (AIC-2), syringic acid (AIC-3), and ursolic acid (UA), at different concentrations (500, 750, and 1000 ppm) were tested on spore germination of Alternaria alternata, Curvularia lunata, Fusarium moniliforme, F. pallidoroseum, and Helminthosporium sp. Significant growth inhibition (P < 0.001) was observed against all fungi except A. alternata which was found to be resistant to AIC-3. Further, two separate sets of field experiments involving spraying of water and F. moniliforme suspension over chemicals treated (1000 ppm) sorghum panicles were done. The levels of protection varied with different treatments which were graded using a standard 1 - 9 rating scale. The Fusarium-challenged panicles (FCP) showed lesser susceptibility and decreased the rate of infection of grain mold (grade 7.0), compared to simple UA, AIC-2, and AIC-1 treatments (7.4, 7.6, and 8.0 grade, resp.). The HPLC quantification of differentially induced phenolic acids in treated sorghum grains substantiated this effect disclosing the higher accumulation of chlorogenic, vanillic, and salicylic acids in FCP. This might be due to defensive induction of these acids by the plants. Although mold control by examined chemicals were lesser than the standard Tilt (grade 5.9), they were found to be nontoxic to mammalian cells under cytotoxicity assay.

No abstract