Modern systems for multi-hop question answering (QA) typically break questions into a sequence of reasoning steps, termed chain-ofthought (CoT), before arriving at a final answer.Often, multiple chains are sampled and aggregated through a voting mechanism over the final answers, but the intermediate steps themselves are discarded.While such approaches improve performance, they do not consider the relations between intermediate steps across chains and do not provide a unified explanation for the predicted answer.We introduce Multi-Chain Reasoning (MCR), an approach which prompts large language models to meta-reason over multiple chains of thought, rather than aggregate their answers.MCR examines different reasoning chains, mixes information between them and selects the most relevant facts in generating an explanation and predicting the answer.MCR outperforms strong baselines on 7 multi-hop QA datasets.Moreover, our analysis reveals that MCR explanations exhibit high quality, enabling humans to verify its answers.

The primary paradigm for multi-task training in natural language processing is to represent the input with a shared pre-trained language model, and add a small, thin network (head) per task.Given an input, a target head is the head that is selected for outputting the final prediction.In this work, we examine the behaviour of non-target heads, that is, the output of heads when given input that belongs to a different task than the one they were trained for.We find that non-target heads exhibit emergent behaviour, which may either explain the target task, or generalize beyond their original task.For example, in a numerical reasoning task, a span extraction head extracts from the input the arguments to a computation that results in a number generated by a target generative head.In addition, a summarization head that is trained with a target question answering head, outputs query-based summaries when given a question and a context from which the answer is to be extracted.This emergent behaviour suggests that multi-task training leads to nontrivial extrapolation of skills, which can be harnessed for interpretability and generalization.

We consider the problem of multiuser detection for randomly spread direct-sequence (DS) code-division multiple access (CDMA) over flat fading channels. The analysis focuses on the case of many users, and large spreading sequences such that their ratio, which is the system load, is kept fixed. Spectral efficiency of practical linear detectors such as match-filter and decorrelator employing successive interference cancellation (SIC) at the receiver is derived. This is used to extend the notion of strongest users detectors for SIC receivers. The strongest users detectors system design relies on an outage approach where each user transmits in a single layer (fixed rate), and only users experiencing good channel conditions may be reliably decoded, while the other users are in outage, i.e., not decoded. In this scheme, iterative SIC decoding is studied, and it is shown that for equal power users, the optimal rate allocation, for maximizing the expected spectral efficiency, is equal rates for all users. This outage approach analysis is extended for multilayer coding broadcast approach per user. The expected sum-rate, under iterative decoding with linear multiuser detectors, is optimized, and the optimal layering power distribution is obtained. For small system loads, the achievable spectral efficiency with the continuous broadcast approach and a linear matched filter detector exhibits significant gains over the single layer coding approach.