Efficient Offline Reinforcement Learning: The Critic is Critical

Recent work has demonstrated both benefits and limitations from using supervised approaches (without temporal-difference learning) for offline reinforcement learning. While off-policy reinforcement learning provides a promising approach for improving performance beyond supervised approaches, we observe that training is often inefficient and unstable due to temporal difference bootstrapping. In this paper we propose a best-of-both approach by first learning the behavior policy and critic with supervised learning, before improving with off-policy reinforcement learning. Specifically, we demonstrate improved efficiency by pre-training with a supervised Monte-Carlo value-error, making use of commonly neglected downstream information from the provided offline trajectories. We find that we are able to more than halve the training time of the considered offline algorithms on standard benchmarks, and surprisingly also achieve greater stability. We further build on the importance of having consistent policy and value functions to propose novel hybrid algorithms, TD3+BC+CQL and EDAC+BC, that regularize both the actor and the critic towards the behavior policy. This helps to more reliably improve on the behavior policy when learning from limited human demonstrations. Code is available at https://github.com/AdamJelley/EfficientOfflineRL

einspace: Searching for Neural Architectures from Fundamental Operations

Neural architecture search (NAS) finds high performing networks for a given task. Yet the results of NAS are fairly prosaic; they did not e.g. create a shift from convolutional structures to transformers. This is not least because the search spaces in NAS often aren't diverse enough to include such transformations a priori. Instead, for NAS to provide greater potential for fundamental design shifts, we need a novel expressive search space design which is built from more fundamental operations. To this end, we introduce einspace, a search space based on a parameterised probabilistic context-free grammar. Our space is versatile, supporting architectures of various sizes and complexities, while also containing diverse network operations which allow it to model convolutions, attention components and more. It contains many existing competitive architectures, and provides flexibility for discovering new ones. Using this search space, we perform experiments to find novel architectures as well as improvements on existing ones on the diverse Unseen NAS datasets. We show that competitive architectures can be obtained by searching from scratch, and we consistently find large improvements when initialising the search with strong baselines. We believe that this work is an important advancement towards a transformative NAS paradigm where search space expressivity and strategic search initialisation play key roles.

Domain-specific augmentations with resolution agnostic self-attention mechanism improves choroid segmentation in optical coherence tomography images

The choroid is a key vascular layer of the eye, supplying oxygen to the retinal photoreceptors. Non-invasive enhanced depth imaging optical coherence tomography (EDI-OCT) has recently improved access and visualisation of the choroid, making it an exciting frontier for discovering novel vascular biomarkers in ophthalmology and wider systemic health. However, current methods to measure the choroid often require use of multiple, independent semi-automatic and deep learning-based algorithms which are not made open-source. Previously, Choroidalyzer -- an open-source, fully automatic deep learning method trained on 5,600 OCT B-scans from 385 eyes -- was developed to fully segment and quantify the choroid in EDI-OCT images, thus addressing these issues. Using the same dataset, we propose a Robust, Resolution-agnostic and Efficient Attention-based network for CHoroid segmentation (REACH). REACHNet leverages multi-resolution training with domain-specific data augmentation to promote generalisation, and uses a lightweight architecture with resolution-agnostic self-attention which is not only faster than Choroidalyzer's previous network (4 images/s vs. 2.75 images/s on a standard laptop CPU), but has greater performance for segmenting the choroid region, vessels and fovea (Dice coefficient for region 0.9769 vs. 0.9749, vessels 0.8612 vs. 0.8192 and fovea 0.8243 vs. 0.3783) due to its improved hyperparameter configuration and model training pipeline. REACHNet can be used with Choroidalyzer as a drop-in replacement for the original model and will be made available upon publication.

Diffusion for World Modeling: Visual Details Matter in Atari

World models constitute a promising approach for training reinforcement learning agents in a safe and sample-efficient manner. Recent world models predominantly operate on sequences of discrete latent variables to model environment dynamics. However, this compression into a compact discrete representation may ignore visual details that are important for reinforcement learning. Concurrently, diffusion models have become a dominant approach for image generation, challenging well-established methods modeling discrete latents. Motivated by this paradigm shift, we introduce DIAMOND (DIffusion As a Model Of eNvironment Dreams), a reinforcement learning agent trained in a diffusion world model. We analyze the key design choices that are required to make diffusion suitable for world modeling, and demonstrate how improved visual details can lead to improved agent performance. DIAMOND achieves a mean human normalized score of 1.46 on the competitive Atari 100k benchmark; a new best for agents trained entirely within a world model. To foster future research on diffusion for world modeling, we release our code, agents and playable world models at https://github.com/eloialonso/diamond.

LLM-Personalize: Aligning LLM Planners with Human Preferences via Reinforced Self-Training for Housekeeping Robots

Large language models (LLMs) have shown significant potential for robotics applications, particularly task planning, by harnessing their language comprehension and text generation capabilities. However, in applications such as household robotics, a critical gap remains in the personalization of these models to individual user preferences. We introduce LLM-Personalize, a novel framework with an optimization pipeline designed to personalize LLM planners for household robotics. Our LLM-Personalize framework features an LLM planner that performs iterative planning in multi-room, partially-observable household scenarios, making use of a scene graph constructed with local observations. The generated plan consists of a sequence of high-level actions which are subsequently executed by a controller. Central to our approach is the optimization pipeline, which combines imitation learning and iterative self-training to personalize the LLM planner. In particular, the imitation learning phase performs initial LLM alignment from demonstrations, and bootstraps the model to facilitate effective iterative self-training, which further explores and aligns the model to user preferences. We evaluate LLM-Personalize on Housekeep, a challenging simulated real-world 3D benchmark for household rearrangements, and show that LLM-Personalize achieves more than a 30 percent increase in success rate over existing LLM planners, showcasing significantly improved alignment with human preferences. Project page: https://donggehan.github.io/projectllmpersonalize/.

Hyperparameter Selection in Continual Learning

In continual learning (CL) -- where a learner trains on a stream of data -- standard hyperparameter optimisation (HPO) cannot be applied, as a learner does not have access to all of the data at the same time. This has prompted the development of CL-specific HPO frameworks. The most popular way to tune hyperparameters in CL is to repeatedly train over the whole data stream with different hyperparameter settings. However, this end-of-training HPO is unrealistic as in practice a learner can only see the stream once. Hence, there is an open question: what HPO framework should a practitioner use for a CL problem in reality? This paper answers this question by evaluating several realistic HPO frameworks. We find that all the HPO frameworks considered, including end-of-training HPO, perform similarly. We therefore advocate using the realistic and most computationally efficient method: fitting the hyperparameters on the first task and then fixing them throughout training.

Choroidalyzer: An open-source, end-to-end pipeline for choroidal analysis in optical coherence tomography

Purpose: To develop Choroidalyzer, an open-source, end-to-end pipeline for segmenting the choroid region, vessels, and fovea, and deriving choroidal thickness, area, and vascular index. Methods: We used 5,600 OCT B-scans (233 subjects, 6 systemic disease cohorts, 3 device types, 2 manufacturers). To generate region and vessel ground-truths, we used state-of-the-art automatic methods following manual correction of inaccurate segmentations, with foveal positions manually annotated. We trained a U-Net deep-learning model to detect the region, vessels, and fovea to calculate choroid thickness, area, and vascular index in a fovea-centred region of interest. We analysed segmentation agreement (AUC, Dice) and choroid metrics agreement (Pearson, Spearman, mean absolute error (MAE)) in internal and external test sets. We compared Choroidalyzer to two manual graders on a small subset of external test images and examined cases of high error. Results: Choroidalyzer took 0.299 seconds per image on a standard laptop and achieved excellent region (Dice: internal 0.9789, external 0.9749), very good vessel segmentation performance (Dice: internal 0.8817, external 0.8703) and excellent fovea location prediction (MAE: internal 3.9 pixels, external 3.4 pixels). For thickness, area, and vascular index, Pearson correlations were 0.9754, 0.9815, and 0.8285 (internal) / 0.9831, 0.9779, 0.7948 (external), respectively (all p<0.0001). Choroidalyzer's agreement with graders was comparable to the inter-grader agreement across all metrics. Conclusions: Choroidalyzer is an open-source, end-to-end pipeline that accurately segments the choroid and reliably extracts thickness, area, and vascular index. Especially choroidal vessel segmentation is a difficult and subjective task, and fully-automatic methods like Choroidalyzer could provide objectivity and standardisation.

DLAS: An Exploration and Assessment of the Deep Learning Acceleration Stack

Deep Neural Networks (DNNs) are extremely computationally demanding, which presents a large barrier to their deployment on resource-constrained devices. Since such devices are where many emerging deep learning applications lie (e.g., drones, vision-based medical technology), significant bodies of work from both the machine learning and systems communities have attempted to provide optimizations to accelerate DNNs. To help unify these two perspectives, in this paper we combine machine learning and systems techniques within the Deep Learning Acceleration Stack (DLAS), and demonstrate how these layers can be tightly dependent on each other with an across-stack perturbation study. We evaluate the impact on accuracy and inference time when varying different parameters of DLAS across two datasets, seven popular DNN architectures, four DNN compression techniques, three algorithmic primitives with sparse and dense variants, untuned and auto-scheduled code generation, and four hardware platforms. Our evaluation highlights how perturbations across DLAS parameters can cause significant variation and across-stack interactions. The highest level observation from our evaluation is that the model size, accuracy, and inference time are not guaranteed to be correlated. Overall we make 13 key observations, including that speedups provided by compression techniques are very hardware dependent, and that compiler auto-tuning can significantly alter what the best algorithm to use for a given configuration is. With DLAS, we aim to provide a reference framework to aid machine learning and systems practitioners in reasoning about the context in which their respective DNN acceleration solutions exist in. With our evaluation strongly motivating the need for co-design, we believe that DLAS can be a valuable concept for exploring the next generation of co-designed accelerated deep learning solutions.

Planning to Go Out-of-Distribution in Offline-to-Online Reinforcement Learning

Offline pretraining with a static dataset followed by online fine-tuning (offline-to-online, or OtO) is a paradigm that is well matched to a real-world RL deployment process: in few real settings would one deploy an offline policy with no test runs and tuning. In this scenario, we aim to find the best-performing policy within a limited budget of online interactions. Previous work in the OtO setting has focused on correcting for bias introduced by the policy-constraint mechanisms of offline RL algorithms. Such constraints keep the learned policy close to the behavior policy that collected the dataset, but this unnecessarily limits policy performance if the behavior policy is far from optimal. Instead, we forgo policy constraints and frame OtO RL as an exploration problem: we must maximize the benefit of the online data-collection. We study major online RL exploration paradigms, adapting them to work well with the OtO setting. These adapted methods contribute several strong baselines. Also, we introduce an algorithm for planning to go out of distribution (PTGOOD), which targets online exploration in relatively high-reward regions of the state-action space unlikely to be visited by the behavior policy. By leveraging concepts from the Conditional Entropy Bottleneck, PTGOOD encourages data collected online to provide new information relevant to improving the final deployment policy. In that way the limited interaction budget is used effectively. We show that PTGOOD significantly improves agent returns during online fine-tuning and finds the optimal policy in as few as 10k online steps in Walker and in as few as 50k in complex control tasks like Humanoid. Also, we find that PTGOOD avoids the suboptimal policy convergence that many of our baselines exhibit in several environments.

Chunking: Forgetting Matters in Continual Learning even without Changing Tasks

Work on continual learning (CL) has largely focused on the problems arising from the dynamically-changing data distribution. However, CL can be decomposed into two sub-problems: (a) shifts in the data distribution, and (b) dealing with the fact that the data is split into chunks and so only a part of the data is available to be trained on at any point in time. In this work, we look at the latter sub-problem -- the chunking of data -- and note that previous analysis of chunking in the CL literature is sparse. We show that chunking is an important part of CL, accounting for around half of the performance drop from offline learning in our experiments. Furthermore, our results reveal that current CL algorithms do not address the chunking sub-problem, only performing as well as plain SGD training when there is no shift in the data distribution. We analyse why performance drops when learning occurs on chunks of data, and find that forgetting, which is often seen to be a problem due to distribution shift, still arises and is a significant problem. Motivated by an analysis of the linear case, we show that per-chunk weight averaging improves performance in the chunking setting and that this performance transfers to the full CL setting. Hence, we argue that work on chunking can help advance CL in general.