We introduce a new family of toy problems that combine features of linear-regression-style continuous in-context learning (ICL) with discrete associative recall. We pretrain transformer models on sample traces from this toy, specifically symbolically-labeled interleaved state observations from randomly drawn linear deterministic dynamical systems. We study if the transformer models can recall the state of a sequence previously seen in its context when prompted to do so with the corresponding in-context label. Taking a closer look at this task, it becomes clear that the model must perform two functions: (1) identify which system's state should be recalled and apply that system to its last seen state, and (2) continuing to apply the correct system to predict the subsequent states. Training dynamics reveal that the first capability emerges well into a model's training. Surprisingly, the second capability, of continuing the prediction of a resumed sequence, develops much earlier. Via out-of-distribution experiments, and a mechanistic analysis on model weights via edge pruning, we find that next-token prediction for this toy problem involves at least two separate mechanisms. One mechanism uses the discrete symbolic labels to do the associative recall required to predict the start of a resumption of a previously seen sequence. The second mechanism, which is largely agnostic to the discrete symbolic labels, performs a "Bayesian-style" prediction based on the previous token and the context. These two mechanisms have different learning dynamics. To confirm that this multi-mechanism (manifesting as separate phase transitions) phenomenon is not just an artifact of our toy setting, we used OLMo training checkpoints on an ICL translation task to see a similar phenomenon: a decisive gap in the emergence of first-task-token performance vs second-task-token performance.

Speech emotion recognition (SER) is vital for obtaining emotional intelligence and understanding the contextual meaning of speech. Variations of consonant-vowel (CV) phonemic boundaries can enrich acoustic context with linguistic cues, which impacts SER. In practice, speech emotions are treated as single labels over an acoustic segment for a given time duration. However, phone boundaries within speech are not discrete events, therefore the perceived emotion state should also be distributed over potentially continuous time-windows. This research explores the implication of acoustic context and phone boundaries on local markers for SER using an attention-based approach. The benefits of using a distributed approach to speech emotion understanding are supported by the results of cross-corpora analysis experiments. Experiments where phones and words are mapped to the attention vectors along with the fundamental frequency to observe the overlapping distributions and thereby the relationship between acoustic context and emotion. This work aims to bridge psycholinguistic theory research with computational modelling for SER.

Off-policy evaluation (OPE) attempts to predict the performance of counterfactual policies using log data from a different policy. We extend its applicability by developing an OPE method for a class of both full support and deficient support logging policies in contextual-bandit settings. This class includes deterministic bandit (such as Upper Confidence Bound) as well as deterministic decision-making based on supervised and unsupervised learning. We prove that our method's prediction converges in probability to the true performance of a counterfactual policy as the sample size increases. We validate our method with experiments on partly and entirely deterministic logging policies. Finally, we apply it to evaluate coupon targeting policies by a major online platform and show how to improve the existing policy.

To facilitate learning, the learning rate is dynamically tuned during training. Training data are randomly partitioned into a "Train Segment" and a "Test Segment." During training on the Train Segment, the network's predictive performance is continually monitored with respect to the Test Segment to help avoid overtraining. To help find the best model and to avoid unsatisfactory local minima, a large number (almost 200) number of candidate models are built and tested for a variety of network configurations and initial weight vectors. Model selection among the candidate models is based solely on predictive performance with respect to independent test data in the Test Segment.

We trained on each acoustic feature a full covariance, genderindependent UBM model with 2048 Gaussians followed by a 600-dimensional i-vector extractor to establish our MFCCand PLP-based i-vector systems. The unlabeled set of development data was used in the training of both the UBM and the i-vector extractor. The open-source Kaldi software has been used for all these processing steps [20]. It has been shown that successive acoustic observation vectors tend to be highly correlated. This may be problematic for maximum a posteriori (MAP) estimation of i-vectors. To investigating this issue, scaling the zero and first order Baum-Welch statistics before presenting them to the i-vector extractor has been proposed. It turns out that a scale factor of 0.33 gives a slight edge, resulting in a better decision cost function [10]. This scaling factor has been performed in training the i-vector extractor as well as in the testing.