Cue-approach training has been shown to effectively shift choices for snack food items by associating a cued button-press motor response to particular food items. Furthermore, attention was biased toward previously cued items, even when the cued item is not chosen for real consumption during a choice phase. However, the exact mechanism by which preferences shift during cue-approach training is not entirely clear. In three experiments, we shed light on the possible underlying mechanisms at play during this novel paradigm: (1) Uncued, wholly predictable motor responses paired with particular food items were not sufficient to elicit a preference shift; (2) Cueing motor responses early – concurrently with food item onset – and thus eliminating the need for heightened top–down attention to the food stimulus in preparation for a motor response also eliminated the shift in food preferences. This finding reinforces our hypothesis that heightened attention at behaviorally relevant points in time is key to changing choice behavior in the cue-approach task; (3) Crucially, indicating choice using eye movements rather than manual button presses preserves the effect, thus demonstrating that the shift in preferences is not governed by a learned motor response but more likely via modulation of subjective value in higher associative regions, consistent with previous neuroimaging results. Cue-approach training drives attention at behaviorally relevant points in time to modulate the subjective value of individual items, providing a mechanism for behavior change that does not rely on external reinforcement and that holds great promise for developing real world behavioral interventions.

This paper presents component techniques essential for converting executables to a high-level intermediate representation (IR) of an existing compiler. The compiler IR is then employed for three distinct applications: binary rewriting using the compiler's binary back-end, vulnerability detection using source-level symbolic execution, and source-code recovery using the compiler's C backend. Our techniques enable complex high-level transformations not possible in existing binary systems, address a major challenge of inputderived memory addresses in symbolic execution and are the first to enable recovery of a fully functional source-code.We present techniques to segment the flat address space in an executable containing undifferentiated blocks of memory. We demonstrate the inadequacy of existing variable identification methods for their promotion to symbols and present our methods for symbol promotion. We also present methods to convert the physically addressed stack in an executable (with a stack pointer) to an abstract stack (without a stack pointer). Our methods do not use symbolic, relocation, or debug information since these are usually absent in deployed executables.We have integrated our techniques with a prototype x86 binary framework called SecondWrite that uses LLVM as IR. The robustness of the framework is demonstrated by handling executables totaling more than a million lines of source-code, produced by two different compilers (gcc and Microsoft Visual Studio compiler), three languages (C, C++, and Fortran), two operating systems (Windows and Linux) and a real world program (Apache server).

An important yet unresolved question is whether or not metacognition consists of domain-general or domain-specific mechanisms. While most studies on this topic suggest a dissociation between metacognitive abilities at the neural level, there are inconsistent reports at the behavioral level. Specifically, while McCurdy et al. (2013) found a positive correlation between metacognitive efficiency for visual perception and memory, such correlation was not observed in Baird et al. (2013). One possible explanation for this discrepancy is that the former included two-alternative-forced choice (2AFC) judgments in both their visual and memory tasks, whereas the latter used 2AFC for one task and yes/no (YN) judgments for the other. To test the effect of task on cross-domain association in metacognitive efficiency, we conducted two online experiments to mirror McCurdy et al. (2013) and Baird et al. (2013) with considerable statistical power (n = 100), and replicated the main findings of both studies. The results suggest that the use of task could affect cross-domain association in metacognitive efficiency. In the third experiment with the same sample size, we used YN judgments for both tasks and did not find a significant cross-domain correlation in metacognitive efficiency. This suggests that the cross-domain correlation found in McCurdy et al. (2013) was not simply due to the same task being used for both domains, and the absence of cross-domain correlation in Baird et al. (2013) might be due to the use of YN judgments. Our results highlight the importance of avoiding confusion between 2AFC and YN judgments in behavioral tasks for metacognitive research, which is a common problem in many behavioral studies.

Recent experiments demonstrate that invisible stimulus features can induce binocular rivalry, indicating the phenomenon may be caused by differences in perceptual signal strength rather than conscious selection processes. Here, we clarify binocular rivalry’s role in consciousness research by highlighting a critical difference between two distinct types of visual awareness.

An important yet unresolved question is whether or not metacognition consists of domain-general or domain-specific mechanisms. While most studies on this topic suggest a dissociation between metacognitive abilities at the neural level, there are conflicting reports at the behavioral level. Our results underscore the need to avoid conflating 2AFC and YN judgments, which is a common problem.peer-reviewed)