{"total":19,"items":[{"citing_arxiv_id":"2606.31282","ref_index":44,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Revisiting the Volume Hypothesis","primary_cat":"cs.LG","submitted_at":"2026-06-30T07:58:19+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"The generalization advantage of SGD over random sampling diminishes with growing training set size in binary networks, as measured by joint density of states over train and test accuracy.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.06959","ref_index":184,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Locally Near Optimal Piecewise Linear Regression in High Dimensions via Difference of Max-Affine Functions","primary_cat":"stat.ML","submitted_at":"2026-05-07T21:23:54+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"ABGD parametrizes piecewise linear functions as difference of max-affine functions and converges linearly to an epsilon-accurate solution with O(d max(sigma/epsilon,1)^2) samples under sub-Gaussian noise, which is minimax optimal up to logs.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.08183","ref_index":36,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Sparsity Hurts: Simple Linear Adapter Can Boost Generalized Category Discovery","primary_cat":"cs.CV","submitted_at":"2026-05-05T10:14:30+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"LAGCD inserts residual linear adapters into each ViT block plus a distribution alignment loss to improve generalized category discovery by increasing model flexibility while reducing bias between seen and novel classes.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"component in adapters: the intrinsic ReLU activation func- tion, which introduces non-linearity and sparsity to enhance performance [22]. Rather than emphasizing the tuning of the bottleneck dimension ˆdand the scaling factors a in adapters, we instead examine how activation functions influence model adaptation. Based on the evaluation of ReLU and its variants in deep neural networks, Xu et al. [36] demon- strated that introducing a non-zero slope for negative inputs in ReLU consistently improves performance. Intuitively, ReLU-induced sparsity through zeroing negative values may discard important information during feature propaga- tion, potentially constraining model adaptation. Motivated by this consideration, we perform a systematic analysis of"},{"citing_arxiv_id":"2604.21119","ref_index":79,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Materialistic RIR: Material Conditioned Realistic RIR Generation","primary_cat":"cs.CV","submitted_at":"2026-04-22T22:04:35+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A two-module neural model disentangles spatial layout from material properties to generate controllable and more realistic room impulse responses, reporting gains of up to 16% on acoustic metrics and 70% on material metrics plus better human ratings.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.16426","ref_index":38,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Functional Similarity Metric for Neural Networks: Overcoming Parametric Ambiguity via Activation Region Analysis","primary_cat":"cs.LG","submitted_at":"2026-04-04T17:50:07+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A functional similarity metric for ReLU networks uses normalized activation region signatures and MinHash to overcome parametric symmetries like neuron permutation and scaling.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2510.04378","ref_index":13,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Score-based Greedy Search for Structure Identification of Partially Observed Linear Causal Models","primary_cat":"cs.LG","submitted_at":"2025-10-05T21:50:17+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Introduces Generalized N Factor Model and LGES algorithm that identifies true causal structure including latents up to Markov equivalence class via score-based greedy search.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2509.22562","ref_index":26,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Activation Function Design Sustains Plasticity in Continual Learning","primary_cat":"cs.LG","submitted_at":"2025-09-26T16:41:47+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Smooth-Leaky and Randomized Smooth-Leaky activations mitigate loss of plasticity in continual learning by targeting negative-branch shape and saturation behavior.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2412.20091","ref_index":101,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Gamma-Ray Burst Light Curve Reconstruction: A Comparative Machine and Deep Learning Analysis","primary_cat":"astro-ph.HE","submitted_at":"2024-12-28T09:20:33+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"MLP and Attention U-Net outperform other models in reconstructing GRB light curves on 521 events, cutting plateau parameter uncertainties by 37-41% versus the Willingale baseline while achieving low MSE.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2410.09355","ref_index":98,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"On Divergence Measures for Training GFlowNets","primary_cat":"cs.LG","submitted_at":"2024-10-12T03:46:52+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Introduces statistically efficient estimators for Renyi-α, Tsallis-α, reverse and forward KL divergences with REINFORCE and score-matching control variates for faster GFlowNet training.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2309.09550","ref_index":67,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Adaptive Reorganization of Neural Pathways for Continual Learning with Spiking Neural Networks","primary_cat":"cs.NE","submitted_at":"2023-09-18T07:56:40+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"SOR-SNN employs Self-Organizing Regulation networks to reorganize a single SNN into sparse pathways, achieving better performance, energy efficiency, memory use, backward transfer, and self-repair on continual learning tasks including CIFAR100 and ImageNet.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1907.08070","ref_index":30,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Discriminative Embedding Autoencoder with a Regressor Feedback for Zero-Shot Learning","primary_cat":"cs.CV","submitted_at":"2019-07-18T14:19:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"A new autoencoder model with margin-based discriminative embeddings and regressor feedback outperforms prior zero-shot learning methods on SUN, CUB, AWA1 and AWA2, with larger gains in generalized ZSL.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1907.07772","ref_index":1,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Modern CNNs for IoT Based Farms","primary_cat":"cs.CY","submitted_at":"2019-07-15T19:28:46+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":2.0,"formal_verification":"none","one_line_summary":"A survey of state-of-the-art CNN architectures for agricultural IoT applications that proposes a tailored classification taxonomy and reviews existing research to guide architecture selection.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1907.05006","ref_index":30,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Two-stream Spatiotemporal Feature for Video QA Task","primary_cat":"cs.CV","submitted_at":"2019-07-11T05:51:39+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"A two-stream spatiotemporal feature extractor with squeeze-and-excitation and attention-based context matching improves text-only video QA on TVQA but shows limitations with visual features.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1907.02994","ref_index":100,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Deep learning in ultrasound imaging","primary_cat":"eess.SP","submitted_at":"2019-07-05T18:39:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":2.0,"formal_verification":"none","one_line_summary":"A review outlining deep learning strategies for adaptive beamforming, spectral Doppler, compressive color Doppler encodings, and structured signal recovery in ultrasound.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"output stride of 2, followed by 2× 2 nearest-neighbour up- sampling. The last block consists of two convolution layers, of which the second again has an output stride of 2, preceding another 5x5 convolution which maps the feature space to a single-channel image through a linear activation function. All other activation functions in the network were leaky rectified linear units [100]. The full deep encoder-decoder network (see Fig. 6a) effectively scales the input image dimensions up by a factor 8, and provides a powerful model that has the capacity to learn the sparse decoding problem, while yielding simultaneous denoising through the compact latent space. The network is trained on simulations of contrast-enhanced ultrasound acquisitions, using an estimate of the real system"},{"citing_arxiv_id":"1907.02959","ref_index":36,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"High-throughput Onboard Hyperspectral Image Compression with Ground-based CNN Reconstruction","primary_cat":"eess.IV","submitted_at":"2019-07-05T17:59:25+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Prequantization-based lossless predictive compression onboard hyperspectral images with CNN ground reconstruction recovers the entire SNR drop at 2 bpp.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1907.00068","ref_index":13,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"On Reducing Negative Jacobian Determinant of the Deformation Predicted by Deep Registration Networks","primary_cat":"cs.CV","submitted_at":"2019-06-28T20:42:12+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Two training mechanisms for unsupervised deep registration networks reduce the number of locations with negative Jacobian determinants in predicted deformations.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1906.09610","ref_index":40,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Improving Description-based Person Re-identification by Multi-granularity Image-text Alignments","primary_cat":"cs.CV","submitted_at":"2019-06-23T17:06:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"The MIA model with GC, RGA, and BFM modules achieves state-of-the-art performance on the CUHK-PEDES dataset for description-based person re-identification.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1710.05941","ref_index":19,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Searching for Activation Functions","primary_cat":"cs.NE","submitted_at":"2017-10-16T18:05:45+00:00","verdict":"CONDITIONAL","verdict_confidence":"MODERATE","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Automated search discovers Swish activation f(x) = x * sigmoid(βx) that improves top-1 ImageNet accuracy over ReLU by 0.9% on Mobile NASNet-A and 0.6% on Inception-ResNet-v2.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1511.06434","ref_index":20,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks","primary_cat":"cs.LG","submitted_at":"2015-11-19T22:50:32+00:00","verdict":"ACCEPT","verdict_confidence":"MODERATE","novelty_score":8.0,"formal_verification":"none","one_line_summary":"DCGANs with architectural constraints learn a hierarchy of representations from object parts to scenes in both generator and discriminator across image datasets.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null}],"limit":50,"offset":0}