Author name cluster

Kangil Kim

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

5 papers

2 author rows

NeurIPS Conference 2025 Conference Paper

Preference Distillation via Value based Reinforcement Learning

Minchan Kwon
Junwon Ko
Kangil kim
Junmo Kim

Direct Preference Optimization (DPO) is a powerful paradigm to align language models with human preferences using pairwise comparisons. However, its binary win-or-loss supervision often proves insufficient for training small models with limited capacity. Prior works attempt to distill information from large teacher models using behavior cloning or KL divergence. These methods often focus on mimicking current behavior and overlook distilling reward modeling. To address this issue, we propose \textit{Teacher Value-based Knowledge Distillation} (TVKD), which introduces an auxiliary reward from the value function of the teacher model to provide a soft guide. This auxiliary reward is formulated to satisfy potential-based reward shaping, ensuring that the global reward structure and optimal policy of DPO are preserved. TVKD can be integrated into the standard DPO training framework and does not require additional rollouts. Our experimental results show that TVKD consistently improves performance across various benchmarks and model sizes.

PDF Details

TMLR Journal 2024 Journal Article

CFASL: Composite Factor-Aligned Symmetry Learning for Disentanglement in Variational AutoEncoder

Hee-Jun Jung
Jaehyoung Jeong
Kangil kim

Symmetries of input and latent vectors have provided valuable insights for disentanglement learning in VAEs. However, only a few works were proposed as an unsupervised method, and even these works require known factor information in training data. We propose a novel method, Composite Factor-Aligned Symmetry Learning (CFASL), which is integrated into VAEs for learning symmetry-based disentanglement in unsupervised learning without any knowledge of the dataset factor information. CFASL incorporates three novel features for learning symmetry-based disentanglement: 1) Injecting inductive bias to align latent vector dimensions to factor-aligned symmetries within an explicit learnable symmetry code-book 2) Learning a composite symmetry to express unknown factors change between two random samples by learning factor-aligned symmetries within the codebook 3) Inducing group equivariant encoder and decoder in training VAEs with the two conditions. In addition, we propose an extended evaluation metric for multi-factor changes in comparison to disentanglement evaluation in VAEs. In quantitative and in-depth qualitative analysis, CFASL demonstrates a significant improvement of disentanglement in single-factor change, and multi-factor change conditions compared to state-of-the-art methods.

PDF Details

ICLR Conference 2024 Conference Paper

Fixed Non-negative Orthogonal Classifier: Inducing Zero-mean Neural Collapse with Feature Dimension Separation

Hoyong Kim
Kangil Kim

Fixed classifiers in neural networks for classification problems have demonstrated cost efficiency and even outperformed learnable classifiers in some popular benchmarks when incorporating orthogonality. Despite these advantages, prior research has yet to investigate the training dynamics of fixed orthogonal classifiers on neural collapse, a recently clarified phenomenon that last-layer features converge to a specific form, called simplex ETF, in training classification models involving the post-zero-error phase. Ensuring this phenomenon is critical for obtaining global optimality in a layer-peeled model, potentially leading to enhanced performance in practice. However, fixed orthogonal classifiers cannot invoke neural collapse due to their geometric limitations. To overcome the limits, we analyze a $\textit{zero-mean neural collapse}$ considering the orthogonality in non-negative Euclidean space. Then, we propose a $\textit{fixed non-negative orthogonal classifier}$ that induces the optimal solution and maximizes the margin of an orthogonal layer-peeled model by satisfying the properties of zero-mean neural collapse. Building on this foundation, we exploit a $\textit{feature dimension separation}$ effect inherent in our classifier for further purposes: (1) enhances softmax masking by mitigating feature interference in continual learning and (2) tackles the limitations of mixup on the hypersphere in imbalanced learning. We conducted comprehensive experiments on various datasets and architectures and demonstrated significant performance improvements.

Details

ICLR Conference 2024 Conference Paper

Label-Focused Inductive Bias over Latent Object Features in Visual Classification

Ilmin Kang
HyounYoung Bae
Kangil Kim

Most neural networks for classification primarily learn features differentiated by input-domain related information such as visual similarity of objects in an image. While this focus is natural behavior, it can inadvertently introduce an inductive bias that conflicts with unseen relations in an implicit output-domain determined by human labeling based on their own world knowledge. Such conflicts can limit generalization of models by potential dominance of the input-domain focused bias in inference. To overcome this limitation without external resources, we introduce Output-Domain focused Biasing (ODB) training strategy that constructs inductive biases on features differentiated by only output labels. It has four steps: 1) it learns intermediate latent object features in an unsupervised manner; 2) it decouples their visual dependencies by assigning new independent embedding parameters; 3) it captures structured features optimized for the original classification task; and 4) it integrates the structured features with the original visual features for the final prediction. We implement the ODB on a vision transformer architecture, and achieved significant improvements on image classification benchmarks. This paper offers a straightforward and effective method to obtain and utilize output-domain focused inductive bias for classification mapping two different domains.

Details

NeurIPS Conference 2022 Conference Paper

Spherization Layer: Representation Using Only Angles

Hoyong Kim
Kangil kim

In neural network literature, angular similarity between feature vectors is frequently used for interpreting or re-using learned representations. However, the inner product in neural networks partially disperses information over the scales and angles of the involved input vectors and weight vectors. Therefore, when using only angular similarity on representations trained with the inner product, information loss occurs in downstream methods, which limits their performance. In this paper, we proposed the $\textit{spherization layer}$ to represent all information on angular similarity. The layer 1) maps the pre-activations of input vectors into the specific range of angles, 2) converts the angular coordinates of the vectors to Cartesian coordinates with an additional dimension, and 3) trains decision boundaries from hyperplanes, without bias parameters, passing through the origin. This approach guarantees that representation learning always occurs on the hyperspherical surface without the loss of any information unlike other projection-based methods. Furthermore, this method can be applied to any network by replacing an existing layer. We validate the functional correctness of the proposed method in a toy task, retention ability in well-known image classification tasks, and effectiveness in word analogy test and few-shot learning. Code is publicly available at https: //github. com/GIST-IRR/spherization_layer

PDF Details