Souvik Kundu

32 papers · 2016–2025 · 11 conferences · across top CS/AI conferences

Achievements

+9 more ↓

🏃 Academic Marathon (9) 🌍 Conference Polyglot (11) 🌉 Interdisciplinary Bridge 🧭 Keyword Pioneer 🐝 Cross-Pollinator (9)

🌈 Renaissance Researcher (7) 🗺️ Taxonomy Completionist (46) 🌉 Interdisciplinary Bridge 🧬 Topic Evolution 👑 Triple Crown ⚡ Prolific Year (7) 🚀 Conference Pioneer 💎 Century Club (32) 🗃️ Keyword Collector (114)

Conferences

ICLR (5) ACL (4) EMNLP (4) ICCV (4) WACV (4) ICML (3) NIPS (3) ECCV (2) COLING (1) INTERSPEECH (1) NAACL (1)

Top co-authors

Massoud Pedram (6) Peter A. Beerel (6) Gourav Datta (4) Peter Anthony Beerel (3) Hwee Tou Ng (3) Zhangyang Wang (3) Qian Lin (2) Yuke Zhang (2) Peter Beerel (2) Akshat Ramachandran (2)

Keywords

model compression (8) neural network pruning (3) adversarial robustness (2) spiking neural network (2) vision transformer (2) reading comprehension (2) conversation history (2) parameter-efficient fine-tuning (2) low-rank adaptation (2) attention mechanism (2) inference optimization (2) knowledge distillation (2) answer extraction (1) question answering (1) model robustness (1) contrastive learning (1) energy efficiency (1) efficient computing (1) accuracy robustness trade-off (1) object detection (1)

Papers

OuroMamba: A Data-Free Quantization Framework for Vision Mamba ICCV 2025 MaskVD: Region Masking for Efficient Video Object Detection WACV 2025 LVLM-Compress-Bench: Benchmarking the Broader Impact of Large Vision-Language Model Compression NAACL 2025 LAMB: A Training-Free Method to Enhance the Long-Context Understanding of SSMs via Attention-Guided Token Filtering ACL 2025 On-the-Fly Adaptive Distillation of Transformer to Dual-State Linear Attention for Long-Context LLM Serving ICML 2025 Mitigating Hallucinations in Vision-Language Models through Image-Guided Head Suppression EMNLP 2025 LAWCAT: Efficient Distillation from Quadratic to Linear Attention with Convolution across Tokens for Long Context Modeling EMNLP 2025 Scaling Long Context Training Data by Long-Distance Referrals ICLR 2025 MambaExtend: A Training-Free Approach to Improve Long Context Extension of Mamba ICLR 2025 LANTERN: Accelerating Visual Autoregressive Models with Relaxed Speculative Decoding ICLR 2025 Fusing Models with Complementary Expertise ICLR 2024 CLAMP-ViT: Contrastive Data-Free Learning for Adaptive Post-Training Quantization of ViTs ECCV 2024 Junk DNA Hypothesis: Pruning Small Pre-Trained Weights $\textitIrreversibly$ and $\textitMonotonically$ Impairs “Difficult" Downstream Tasks in LLMs ICML 2024 AFLoRA: Adaptive Freezing of Low Rank Adaptation in Parameter Efficient Fine-Tuning of Large Models ACL 2024 LaMDA: Large Model Fine-Tuning via Spectrally Decomposed Low-Dimensional Adaptation EMNLP 2024 ShiftAddLLM: Accelerating Pretrained LLMs via Post-Training Multiplication-Less Reparameterization NIPS 2024 GenQ: Quantization in Low Data Regimes with Generative Synthetic Data ECCV 2024 Vision HGNN: An Image is More than a Graph of Nodes ICCV 2023 Don’t just prune by magnitude! Your mask topology is a secret weapon NIPS 2023 SAL-ViT: Towards Latency Efficient Private Inference on ViT using Selective Attention Search with a Learnable Softmax Approximation ICCV 2023 Learning to Linearize Deep Neural Networks for Secure and Efficient Private Inference ICLR 2023 NeRFool: Uncovering the Vulnerability of Generalizable Neural Radiance Fields against Adversarial Perturbations ICML 2023 Self-Attentive Pooling for Efficient Deep Learning WACV 2023 FLOAT: Fast Learnable Once-for-All Adversarial Training for Tunable Trade-Off Between Accuracy and Robustness WACV 2023 Analyzing the Confidentiality of Undistillable Teachers in Knowledge Distillation NIPS 2021 HIRE-SNN: Harnessing the Inherent Robustness of Energy-Efficient Deep Spiking Neural Networks by Training With Crafted Input Noise ICCV 2021 Spike-Thrift: Towards Energy-Efficient Deep Spiking Neural Networks by Limiting Spiking Activity via Attention-Guided Compression WACV 2021 A Co-Attentive Cross-Lingual Neural Model for Dialogue Breakdown Detection COLING 2020 Learning to Identify Follow-Up Questions in Conversational Question Answering ACL 2020 Exploiting Explicit Paths for Multi-hop Reading Comprehension ACL 2019 A Nil-Aware Answer Extraction Framework for Question Answering EMNLP 2018 Incorporating a Generative Front-End Layer to Deep Neural Network for Noise Robust Automatic Speech Recognition INTERSPEECH 2016