Large Language Models (LLMs) have shown remarkable capabilities across tasks, yet they often require additional prompting techniques when facing complex problems. While approaches like self-correction and response selection have emerged as popular solutions, recent studies have shown these methods perform poorly when relying on the LLM itself to provide feedback or selection criteria. We argue this limitation stems from the fact that common LLM post-training procedures lack explicit supervision for discriminative judgment tasks. In this paper, we propose Generative Self-Aggregation (GSA), a novel prompting method that improves answer quality without requiring the model's discriminative capabilities. GSA first samples multiple diverse responses from the LLM, then aggregates them to obtain an improved solution. Unlike previous approaches, our method does not require the LLM to correct errors or compare response quality; instead, it leverages the model's generative abilities to synthesize a new response based on the context of multiple samples. While GSA shares similarities with the self-consistency (SC) approach for response aggregation, SC requires specific verifiable tokens to enable majority voting. In contrast, our approach is more general and can be applied to open-ended tasks. Empirical evaluation demonstrates that GSA effectively improves response quality across various tasks, including mathematical reasoning, knowledge-based problems, and open-ended generation tasks such as code synthesis and conversational responses.

For decades, the cougar (Puma concolor) has been thought of as a loner predator, running across other members of its species only to mate or to fight. But a new study now shows that, contrary to popular belief, cougars have quietly built for themselves a rich, hierarchical society based largely on sharing food--a find that stands to upend scientists' preconceptions about one of the Americas' most iconic big cats. "For more than 60 years of intensive research... we have said that [cougars] are solitary, robotic killing machines," says Mark Elbroch, lead scientist for the Puma Program at Panthera, the global wild cat conservation organization. "Instead, what we have unveiled is a secretive animal with a complex social system completely built on reciprocity. "That flies in the face of everything we ever thought about this animal," he adds. "I thought it was really exciting research--we usually assume that these animals aren't social," adds Justine Smith, a postdoctoral researcher at the ...