Yi Shen

Zifan Wang, Yi Shen, Zachary I. Bell et al.

We consider risk-averse learning in repeated unknown games where the goal of the agents is to minimize their individual risk of incurring significantly high cost. Specifically, the agents use the conditional value at risk (CVaR) as a risk measure and rely on bandit feedback in the form of the cost values of the selected actions at every episode to estimate their CVaR values and update their actions. A major challenge in using bandit feedback to estimate CVaR is that the agents can only access their own cost values, which, however, depend on the actions of all agents. To address this challenge, we propose a new risk-averse learning algorithm with momentum that utilizes the full historical information on the cost values. We show that this algorithm achieves sub-linear regret and matches the best known algorithms in the literature. We provide numerical experiments for a Cournot game that show that our method outperforms existing methods.

Zifan Wang, Yi Shen, Michael M. Zavlanos

We consider an online stochastic game with risk-averse agents whose goal is to learn optimal decisions that minimize the risk of incurring significantly high costs. Specifically, we use the Conditional Value at Risk (CVaR) as a risk measure that the agents can estimate using bandit feedback in the form of the cost values of only their selected actions. Since the distributions of the cost functions depend on the actions of all agents that are generally unobservable, they are themselves unknown and, therefore, the CVaR values of the costs are difficult to compute. To address this challenge, we propose a new online risk-averse learning algorithm that relies on one-point zeroth-order estimation of the CVaR gradients computed using CVaR values that are estimated by appropriately sampling the cost functions. We show that this algorithm achieves sub-linear regret with high probability. We also propose two variants of this algorithm that improve performance. The first variant relies on a new sampling strategy that uses samples from the previous iteration to improve the estimation accuracy of the CVaR values. The second variant employs residual feedback that uses CVaR values from the previous iteration to reduce the variance of the CVaR gradient estimates. We theoretically analyze the convergence properties of these variants and illustrate their performance on an online market problem that we model as a Cournot game.

Yi Shen, Jessilyn Dunn, Michael M. Zavlanos

In this paper, we consider a risk-averse multi-armed bandit (MAB) problem where the goal is to learn a policy that minimizes the risk of low expected return, as opposed to maximizing the expected return itself, which is the objective in the usual approach to risk-neutral MAB. Specifically, we formulate this problem as a transfer learning problem between an expert and a learner agent in the presence of contexts that are only observable by the expert but not by the learner. Thus, such contexts are unobserved confounders (UCs) from the learner's perspective. Given a dataset generated by the expert that excludes the UCs, the goal for the learner is to identify the true minimum-risk arm with fewer online learning steps, while avoiding possible biased decisions due to the presence of UCs in the expert's data.

Chao Yu, Yi Shen, Yue Mao et al.

Hierarchical Text Classification (HTC) is a challenging task where a document can be assigned to multiple hierarchically structured categories within a taxonomy. The majority of prior studies consider HTC as a flat multi-label classification problem, which inevitably leads to "label inconsistency" problem. In this paper, we formulate HTC as a sequence generation task and introduce a sequence-to-tree framework (Seq2Tree) for modeling the hierarchical label structure. Moreover, we design a constrained decoding strategy with dynamic vocabulary to secure the label consistency of the results. Compared with previous works, the proposed approach achieves significant and consistent improvements on three benchmark datasets.

Yuan Zhuang, Yi Shen, Yuexin Bian et al.

Recent studies have shown that combining parameter-efficient fine-tuning (PEFT) with mixture-of-experts (MoE) is an effective strategy for adapting large language models (LLMs) to the downstream tasks. However, most existing approaches rely on conventional TopK routing, which requires careful hyperparameter tuning and assigns a fixed number of experts to each token. In this work, we propose LD-MoLE, a Learnable Dynamic routing mechanism for Mixture of LoRA Experts that enables adaptive, token-dependent, and layer-wise expert allocation. Our method replaces the non-differentiable TopK selection with a differentiable routing function and a closed-form solution. Moreover, our design allows the model to adaptively determine the number of experts to activate for each token at different layers. In addition, we introduce an analytical sparsity control objective to regularize the number of activated experts. Extensive experiments on the Qwen3-1.7B and Llama-3.2-3B models show that LD-MoLE achieves the highest average scores compared to state-of-the-art baselines, across a diverse set of benchmarks. Our method not only achieves superior performance, but also demonstrates the ability to learn token-dependent and layer-wise expert allocation.

Yi Shen, Hanyan Huang

Offline reinforcement learning has received extensive attention from scholars because it avoids the interaction between the agent and the environment by learning a policy through a static dataset. However, general reinforcement learning methods cannot get satisfactory results in offline reinforcement learning due to the out-of-distribution state actions that the dataset cannot cover during training. To solve this problem, the policy regularization method that tries to directly clone policies used in static datasets has received numerous studies due to its simplicity and effectiveness. However, policy constraint methods make the agent choose the corresponding actions in the static dataset. This type of constraint is usually over-conservative, which results in suboptimal policies, especially in low-quality static datasets. In this paper, a hypercube policy regularization framework is proposed, this method alleviates the constraints of policy constraint methods by allowing the agent to explore the actions corresponding to similar states in the static dataset, which increases the effectiveness of algorithms in low-quality datasets. It was also theoretically demonstrated that the hypercube policy regularization framework can effectively improve the performance of original algorithms. In addition, the hypercube policy regularization framework is combined with TD3-BC and Diffusion-QL for experiments on D4RL datasets which are called TD3-BC-C and Diffusion-QL-C. The experimental results of the score demonstrate that TD3-BC-C and Diffusion-QL-C perform better than state-of-the-art algorithms like IQL, CQL, TD3-BC and Diffusion-QL in most D4RL environments in approximate time.

Yi Shen, Hanyan Huang

Offline reinforcement learning learns from a static dataset without interacting with environments, which ensures security and thus owns a good application prospect. However, directly applying naive reinforcement learning algorithm usually fails in an offline environment due to inaccurate Q value approximation caused by out-of-distribution (OOD) state-actions. It is an effective way to solve this problem by penalizing the Q-value of OOD state-actions. Among the methods of punishing OOD state-actions, count-based methods have achieved good results in discrete domains in a simple form. Inspired by it, a novel pseudo-count method for continuous domains called Grid-Mapping Pseudo-Count method (GPC) is proposed by extending the count-based method from discrete to continuous domains. Firstly, the continuous state and action space are mapped to discrete space using Grid-Mapping, then the Q-values of OOD state-actions are constrained through pseudo-count. Secondly, the theoretical proof is given to show that GPC can obtain appropriate uncertainty constraints under fewer assumptions than other pseudo-count methods. Thirdly, GPC is combined with Soft Actor-Critic algorithm (SAC) to get a new algorithm called GPC-SAC. Lastly, experiments on D4RL datasets are given to show that GPC-SAC has better performance and less computational cost than other algorithms that constrain the Q-value.

Lin Yang, Yi Shen, Yue Mao et al.

Emotion recognition in conversation (ERC) aims to detect the emotion label for each utterance. Motivated by recent studies which have proven that feeding training examples in a meaningful order rather than considering them randomly can boost the performance of models, we propose an ERC-oriented hybrid curriculum learning framework. Our framework consists of two curricula: (1) conversation-level curriculum (CC); and (2) utterance-level curriculum (UC). In CC, we construct a difficulty measurer based on "emotion shift" frequency within a conversation, then the conversations are scheduled in an "easy to hard" schema according to the difficulty score returned by the difficulty measurer. For UC, it is implemented from an emotion-similarity perspective, which progressively strengthens the model's ability in identifying the confusing emotions. With the proposed model-agnostic hybrid curriculum learning strategy, we observe significant performance boosts over a wide range of existing ERC models and we are able to achieve new state-of-the-art results on four public ERC datasets.