Deepak Gopinath

Emily Sumner, Deepak E. Gopinath, Laporsha Dees et al.

High-quality curated datasets are essential for training and evaluating AI approaches, but are often lacking in embodied interactive domains where language and physical action are intertwined. In particular, few datasets capture how people acquire motor skills in embodied tasks through verbal instruction over time. To address this gap, we introduce SimCoachCorpus: a unique dataset of race car simulator driving that enables the investigation of rich phenomena during guided and unguided motor skill acquisition. In this dataset, 29 humans were asked to drive in a driving simulator around a race track for approximately ninety minutes. Fifteen participants received one-on-one instruction from a professional performance driving coach, and 14 participants drove without coaching instruction. SimCoachCorpus includes features such as vehicle state and inputs, map (track boundaries and race-line), and cone landmarks. Additionally, these are synchronized with the coach's concurrent verbal feedback and additional terminal feedback at the end of each lap. We also provide high-quality annotations of high-level coaching categories for each concurrent feedback utterance, ratings on students' compliance with coaching advice, and self-reported cognitive load and emotional state of participants (gathered from surveys during the study). The final dataset includes over 20,000 concurrent feedback utterances, over 400 terminal feedback utterances, and over 40 hours of interactive driving data. Our naturalistic interactive dataset can be used to investigate motor learning dynamics, explore linguistic phenomena, and train computational models of teaching and learning. We demonstrate applications of this dataset for in-context learning, imitation learning, and topic modeling. Data is hosted at https://doi.org/10.7910/DVN/W7VTKZ and code is available at https://github.com/ToyotaResearchInstitute/sim_coach_corpus

Tom Gunter, Zirui Wang, Chong Wang et al.

We present foundation language models developed to power Apple Intelligence features, including a ~3 billion parameter model designed to run efficiently on devices and a large server-based language model designed for Private Cloud Compute. These models are designed to perform a wide range of tasks efficiently, accurately, and responsibly. This report describes the model architecture, the data used to train the model, the training process, how the models are optimized for inference, and the evaluation results. We highlight our focus on Responsible AI and how the principles are applied throughout the model development.

Roy Xie, Deepak Gopinath, David Qiu et al.

Search-augmented large language models (LLMs) excel at knowledge-intensive tasks by integrating external retrieval. However, they often over-search -- unnecessarily invoking search tool even when it does not improve response quality, which leads to computational inefficiency and hallucinations by incorporating irrelevant context. In this work, we conduct a systematic evaluation of over-searching across multiple dimensions, including query types, model categories, retrieval conditions, and multi-turn conversations. Our finding shows: (i) search generally improves answer accuracy on answerable queries but harms abstention on unanswerable ones; (ii) over-searching is more pronounced in complex reasoning models and deep research systems, is exacerbated by noisy retrieval, and compounds across turns in multi-turn conversations; and (iii) the composition of retrieved evidence is crucial, as the presence of negative evidence improves abstention. To quantify over-searching, we introduce Tokens Per Correctness (TPC), an evaluation metric that captures the performance-cost trade-off for search-augmented LLMs. Lastly, we investigate mitigation approaches at both the query and retrieval levels and release the OverSearchQA to foster continued research into efficient search-augmented LLMs.

Deepak Gopinath, Guy Rosman, Simon Stent et al.

We propose a computational model to estimate a person's attended awareness of their environment. We define attended awareness to be those parts of a potentially dynamic scene which a person has attended to in recent history and which they are still likely to be physically aware of. Our model takes as input scene information in the form of a video and noisy gaze estimates, and outputs visual saliency, a refined gaze estimate, and an estimate of the person's attended awareness. In order to test our model, we capture a new dataset with a high-precision gaze tracker including 24.5 hours of gaze sequences from 23 subjects attending to videos of driving scenes. The dataset also contains third-party annotations of the subjects' attended awareness based on observations of their scan path. Our results show that our model is able to reasonably estimate attended awareness in a controlled setting, and in the future could potentially be extended to real egocentric driving data to help enable more effective ahead-of-time warnings in safety systems and thereby augment driver performance. We also demonstrate our model's effectiveness on the tasks of saliency, gaze calibration, and denoising, using both our dataset and an existing saliency dataset. We make our model and dataset available at https://github.com/ToyotaResearchInstitute/att-aware/.

Justin Lidard, Haimin Hu, Asher Hancock et al.

As intelligent robots like autonomous vehicles become increasingly deployed in the presence of people, the extent to which these systems should leverage model-based game-theoretic planners versus data-driven policies for safe, interaction-aware motion planning remains an open question. Existing dynamic game formulations assume all agents are task-driven and behave optimally. However, in reality, humans tend to deviate from the decisions prescribed by these models, and their behavior is better approximated under a noisy-rational paradigm. In this work, we investigate a principled methodology to blend a data-driven reference policy with an optimization-based game-theoretic policy. We formulate KLGame, an algorithm for solving non-cooperative dynamic game with Kullback-Leibler (KL) regularization with respect to a general, stochastic, and possibly multi-modal reference policy. Our method incorporates, for each decision maker, a tunable parameter that permits modulation between task-driven and data-driven behaviors. We propose an efficient algorithm for computing multi-modal approximate feedback Nash equilibrium strategies of KLGame in real time. Through a series of simulated and real-world autonomous driving scenarios, we demonstrate that KLGame policies can more effectively incorporate guidance from the reference policy and account for noisily-rational human behaviors versus non-regularized baselines. Website with additional information, videos, and code: https://kl-games.github.io/.

Deepak Edakkattil Gopinath, Xiongyi Cui, Jonathan DeCastro et al.

Learning-based automated coaching systems for complex motor tasks such as high-performance driving remain limited in the ability to be adaptive by their reliance only on local, context-dependent reasoning, failing to account for the long-term temporal nature of student learning and the cumulative impact of repeated teacher-student interactions. In this paper, we propose an imitation learning based computational model for adaptive teaching with a dedicated temporal reasoning module that can reason over the interaction history under low-data regimes. To compensate for limited amounts of interactive training data, and based on the repetitive nature of the teaching process, the model relies on a nearest neighbor retrieval and cross attention prior, reasoning only on a narrowed-down set of semantically similar past interactions with an encoder-decoder based concurrent teaching model. We validate our approach with (i) a novel semi-synthetic closed-loop longitudinal student-teacher interaction dataset based on Waymo Open Motion Dataset and (ii) a small-scale real-world naturalistic simulator race coaching dataset. Our results reveal the consistent advantage of our adaptive teaching model with the nearest neighbor retrieval and cross-attention prior over a non-adaptive baseline as well as a suite of adaptive models that differ in their choice of priors and temporal fusion mechanisms.

Roy Xie, David Qiu, Deepak Gopinath et al.

Long chain-of-thought (CoT) significantly enhances large language models' (LLM) reasoning capabilities. However, the extensive reasoning traces lead to inefficiencies and an increased time-to-first-token (TTFT). We propose a novel training paradigm that uses reinforcement learning (RL) to guide reasoning LLMs to interleave thinking and answering for multi-hop questions. We observe that models inherently possess the ability to perform interleaved reasoning, which can be further enhanced through RL. We introduce a simple yet effective rule-based reward to incentivize correct intermediate steps, which guides the policy model toward correct reasoning paths by leveraging intermediate signals generated during interleaved reasoning. Extensive experiments conducted across five diverse datasets and three RL algorithms (PPO, GRPO, and REINFORCE++) demonstrate consistent improvements over traditional think-answer reasoning, without requiring external tools. Specifically, our approach reduces TTFT by over 80% on average and improves up to 19.3% in Pass@1 accuracy. Furthermore, our method, trained solely on question answering and logical reasoning datasets, exhibits strong generalization ability to complex reasoning datasets such as MATH, GPQA, and MMLU. Additionally, we conduct in-depth analysis to reveal several valuable insights into conditional reward modeling.

Deepak E. Gopinath, Brenna D. Argall

Assistive shared-control robots have the potential to transform the lives of millions of people afflicted with severe motor impairments. The usefulness of shared-control robots typically relies on the underlying autonomy's ability to infer the user's needs and intentions, and the ability to do so unambiguously is often a limiting factor for providing appropriate assistance confidently and accurately. The contributions of this paper are four-fold. First, we introduce the idea of intent disambiguation via control mode selection, and present a mathematical formalism for the same. Second, we develop a control mode selection algorithm which selects the control mode in which the user-initiated motion helps the autonomy to maximally disambiguate user intent. Third, we present a pilot study with eight subjects to evaluate the efficacy of the disambiguation algorithm. Our results suggest that the disambiguation system (a) helps to significantly reduce task effort, as measured by number of button presses, and (b) is of greater utility for more limited control interfaces and more complex tasks. We also observe that (c) subjects demonstrated a wide range of disambiguation request behaviors, with the common thread of concentrating requests early in the execution. As our last contribution, we introduce a novel field-theoretic approach to intent inference inspired by dynamic field theory that works in tandem with the disambiguation scheme.

Juan Pino, Liezl Puzon, Jiatao Gu et al.

For automatic speech translation (AST), end-to-end approaches are outperformed by cascaded models that transcribe with automatic speech recognition (ASR), then translate with machine translation (MT). A major cause of the performance gap is that, while existing AST corpora are small, massive datasets exist for both the ASR and MT subsystems. In this work, we evaluate several data augmentation and pretraining approaches for AST, by comparing all on the same datasets. Simple data augmentation by translating ASR transcripts proves most effective on the English--French augmented LibriSpeech dataset, closing the performance gap from 8.2 to 1.4 BLEU, compared to a very strong cascade that could directly utilize copious ASR and MT data. The same end-to-end approach plus fine-tuning closes the gap on the English--Romanian MuST-C dataset from 6.7 to 3.7 BLEU. In addition to these results, we present practical recommendations for augmentation and pretraining approaches. Finally, we decrease the performance gap to 0.01 BLEU using a Transformer-based architecture.

Mason Bretan, Deepak Gopinath, Philip Mullins et al.

The design and evaluation of a robotic prosthesis for a drummer with a transradial amputation is presented. The principal objective of the prosthesis is to simulate the role fingers play in drumming. This primarily includes controlling the manner in which the drum stick rebounds after initial impact. This is achieved using a DC motor driven by a variable impedance control framework in a shared control system. The user's ability to perform with and control the prosthesis is evaluated using a musical synchronization study. A secondary objective of the prosthesis is to explore the implications of musical expression and human-robotic interaction when a second, completely autonomous, stick is added to the prosthesis. This wearable robotic musician interacts with the user by listening to the music and responding with different rhythms and behaviors. We recount some empirical findings based on the user's experience of performing under such a paradigm.