{"noiseShare":0.4927,"topics":[{"claims":[870,1055,1331,1391,1481,1815,1918,1245,1650],"deltaPp":-3.9136,"id":0,"keywords":["convergence","gradient","stochastic","optimization","descent"],"label":"Gradient-Based Optimization Convergence","labelFull":"Convergence Guarantees for Gradient-Based Optimization","papers":7257,"papersByYear":[471,563,718,801,838,1033,1076,737,1020],"rel":-0.4662,"reviewed":true,"share":[8.3942,7.8774,8.0538,7.7964,7.1265,7.4418,5.688,3.9583,4.4806],"size":12756,"slope":-0.005416},{"claims":[460,546,769,915,1042,1308,1635,1779,2627],"deltaPp":2.2212,"id":1,"keywords":["state-of-the-art","benchmark","outperforms","demonstrate","experiments"],"label":"State-of-the-Art Benchmarking","labelFull":"State-of-the-Art Benchmarking on Standard Datasets","papers":9505,"papersByYear":[408,483,660,794,906,1133,1423,1537,2161],"rel":0.3055,"reviewed":true,"share":[7.2714,6.7581,7.4033,7.7282,7.7047,8.1622,7.5223,8.255,9.4926],"size":11081,"slope":0.002341},{"claims":[465,751,995,1067,1178,1210,1353,1268,1449],"deltaPp":-0.9618,"id":2,"keywords":["adversarial","attacks","robustness","backdoor","defense"],"label":"Adversarial Robustness","labelFull":"Adversarial Robustness: Attacks and Defense Methods","papers":4627,"papersByYear":[244,365,447,486,520,530,644,620,771],"rel":-0.2212,"reviewed":true,"share":[4.3486,5.107,5.014,4.7304,4.4221,3.8182,3.4043,3.3299,3.3868],"size":9736,"slope":-0.002218},{"claims":[240,543,776,810,1018,1374,1452,1124,1043],"deltaPp":0.0526,"id":3,"keywords":["graph","node","networks","neural","classification"],"label":"Graph Neural Networks","labelFull":"Graph Neural Networks: Efficient, Scalable, Robust Models","papers":3622,"papersByYear":[123,236,341,368,423,541,596,483,511],"rel":0.024,"reviewed":true,"share":[2.1921,3.3021,3.825,3.5819,3.5972,3.8974,3.1506,2.5941,2.2447],"size":8380,"slope":-0.000491},{"claims":[354,602,699,861,1024,1231,1305,1083,783],"deltaPp":-1.3516,"id":4,"keywords":["segmentation","semantic","instance","object","image"],"label":"Image Segmentation","labelFull":"Image Segmentation: Semantic, Instance, and Panoptic Tasks","papers":4149,"papersByYear":[187,303,362,444,525,644,676,557,451],"rel":-0.4056,"reviewed":true,"share":[3.3327,4.2395,4.0606,4.3216,4.4647,4.6394,3.5735,2.9916,1.9811],"size":7942,"slope":-0.001634},{"claims":[142,173,285,350,466,610,1780,2177,1909],"deltaPp":4.3371,"id":5,"keywords":["languages","multilingual","fine-tuning","evaluation","framework"],"label":"Multilingual LLM Benchmarking","labelFull":"Benchmarking Multilingual Large Language Models","papers":5474,"papersByYear":[99,128,193,244,323,420,1219,1459,1389],"rel":2.4581,"reviewed":true,"share":[1.7644,1.791,2.1649,2.3749,2.7468,3.0257,6.4439,7.8361,6.1015],"size":7892,"slope":0.007449},{"claims":[680,795,902,830,850,906,856,625,672],"deltaPp":-4.9842,"id":6,"keywords":["networks","neural","relu","deep","recurrent"],"label":"Layerwise Training","labelFull":"Layerwise Training of Deep Neural Networks","papers":4387,"papersByYear":[396,459,516,507,518,558,548,413,472],"rel":-0.7062,"reviewed":true,"share":[7.0576,6.4223,5.788,4.9348,4.4051,4.0199,2.8969,2.2182,2.0734],"size":7116,"slope":-0.006541},{"claims":[43,87,108,136,200,358,827,1502,3720],"deltaPp":8.9146,"id":7,"keywords":["reasoning","mathematical","math","chain-of-thought","benchmarks"],"label":"Chain-of-Thought Efficiency","labelFull":"Chain-of-Thought Efficiency for Large Language Models","papers":3986,"papersByYear":[28,56,69,83,117,213,467,810,2143],"rel":17.8649,"reviewed":true,"share":[0.499,0.7835,0.774,0.8079,0.995,1.5345,2.4687,4.3504,9.4136],"size":6981,"slope":0.008413},{"claims":[197,264,307,474,583,723,1294,1358,1546],"deltaPp":0.955,"id":8,"keywords":["speech","audio","music","acoustic","recognition"],"label":"Audio-Language Models","labelFull":"Audio-Language Models: Pretraining, Evaluation, Adaptation","papers":2403,"papersByYear":[80,116,122,177,217,264,434,451,542],"rel":0.6698,"reviewed":true,"share":[1.4258,1.6231,1.3685,1.7228,1.8454,1.9019,2.2942,2.4223,2.3808],"size":6746,"slope":0.001375},{"claims":[420,469,566,645,725,619,948,1135,1140],"deltaPp":0.284,"id":9,"keywords":["question","answering","qa","reading","knowledge"],"label":"Question Answering","labelFull":"Question Answering: Methods, Datasets, and Evaluation","papers":3572,"papersByYear":[181,197,266,286,331,321,518,673,799],"rel":0.088,"reviewed":true,"share":[3.2258,2.7564,2.9837,2.7837,2.8149,2.3125,2.7383,3.6146,3.5098],"size":6667,"slope":0.000458},{"claims":[335,564,605,665,752,749,977,946,898],"deltaPp":-0.3651,"id":10,"keywords":["dialogue","conversational","intent","response","user"],"label":"End-to-End Dialogue Modeling","labelFull":"End-to-End Dialogue Modeling, State Tracking, Intent","papers":2867,"papersByYear":[142,222,244,268,298,314,438,448,493],"rel":-0.1443,"reviewed":true,"share":[2.5307,3.1062,2.737,2.6085,2.5342,2.2621,2.3154,2.4061,2.1656],"size":6491,"slope":-0.000792},{"claims":[98,131,212,264,413,707,1110,1326,1783],"deltaPp":2.3364,"id":11,"keywords":["protein","molecular","alignment","molecules","design"],"label":"Molecular and Protein Design","labelFull":"Machine Learning for Molecular and Protein Design","papers":2748,"papersByYear":[72,73,112,139,199,309,469,551,824],"rel":1.8208,"reviewed":true,"share":[1.2832,1.0214,1.2563,1.3529,1.6923,2.2261,2.4793,2.9593,3.6196],"size":6044,"slope":0.00308},{"claims":[37,66,85,141,284,582,983,1455,2241],"deltaPp":5.8215,"id":12,"keywords":["vision-language","visual","reasoning","tokens","spatial"],"label":"Vision-Language Model Benchmarking","labelFull":"Benchmarking Reasoning and Robustness of Vision-Language Models","papers":3833,"papersByYear":[30,50,64,101,192,390,649,910,1447],"rel":10.8874,"reviewed":true,"share":[0.5347,0.6996,0.7179,0.9831,1.6328,2.8096,3.4308,4.8875,6.3562],"size":5874,"slope":0.007184},{"claims":[118,235,249,398,479,867,1146,1112,836],"deltaPp":0.8377,"id":13,"keywords":["rendering","3d","radiance","view","gaussian"],"label":"3D Gaussian Splatting","labelFull":"3D Gaussian Splatting for Efficient High-Quality Rendering","papers":3047,"papersByYear":[75,133,141,216,262,485,630,610,495],"rel":0.6267,"reviewed":true,"share":[1.3367,1.8609,1.5816,2.1024,2.2281,3.494,3.3303,3.2762,2.1744],"size":5440,"slope":0.002081},{"claims":[275,300,427,608,647,717,853,517,1042],"deltaPp":0.3887,"id":14,"keywords":["reinforcement","learning","offline","policy","imitation"],"label":"Reinforcement Learning Algorithms","labelFull":"Reinforcement Learning Algorithms and Benchmark Performance","papers":3432,"papersByYear":[166,199,264,351,389,439,516,346,762],"rel":0.1314,"reviewed":true,"share":[2.9585,2.7844,2.9613,3.4164,3.3081,3.1626,2.7277,1.8583,3.3472],"size":5386,"slope":-0.000324},{"claims":[311,381,469,584,637,669,787,735,527],"deltaPp":-1.1909,"id":15,"keywords":["pose","motion","estimation","human","3d"],"label":"3D Pose Estimation","labelFull":"3D Human, Object, and Camera Pose Estimation","papers":2605,"papersByYear":[142,188,223,271,312,358,419,387,305],"rel":-0.4706,"reviewed":true,"share":[2.5307,2.6305,2.5014,2.6377,2.6533,2.5791,2.2149,2.0785,1.3398],"size":5100,"slope":-0.001175},{"claims":[198,300,479,669,803,847,735,502,388],"deltaPp":-1.1641,"id":16,"keywords":["label","contrastive","loss","learning","semi-supervised"],"label":"Label-Efficient Learning","labelFull":"Semi-supervised and Self-supervised Learning with Limited Labels","papers":3288,"papersByYear":[140,218,301,447,517,544,489,329,303],"rel":-0.4666,"reviewed":true,"share":[2.4951,3.0502,3.3763,4.3508,4.3966,3.919,2.585,1.767,1.331],"size":4921,"slope":-0.001753},{"claims":[479,468,596,610,582,459,605,487,385],"deltaPp":-3.161,"id":17,"keywords":["translation","machine","neural","quality","multilingual"],"label":"Neural Machine Translation","labelFull":"Neural Machine Translation: Low-Resource Multilingual Gains","papers":2100,"papersByYear":[220,220,271,274,249,206,263,224,173],"rel":-0.8062,"reviewed":true,"share":[3.9209,3.0782,3.0398,2.6669,2.1175,1.484,1.3903,1.2031,0.7599],"size":4671,"slope":-0.003792},{"claims":[48,71,88,136,203,321,688,1181,1607],"deltaPp":3.9503,"id":18,"keywords":["multimodal","modality","fusion","reasoning","visual"],"label":"Multimodal LLMs","labelFull":"Multimodal LLMs: Training, Alignment, and Evaluation","papers":2751,"papersByYear":[30,48,64,95,139,194,427,733,1021],"rel":7.3879,"reviewed":true,"share":[0.5347,0.6716,0.7179,0.9247,1.1821,1.3976,2.2572,3.9368,4.485],"size":4343,"slope":0.004858},{"claims":[76,95,120,208,299,341,668,1117,1390],"deltaPp":2.7453,"id":19,"keywords":["retrieval","query","generation","dense","information"],"label":"Retrieval-Augmented Generation","labelFull":"Retrieval-Augmented Generation for Knowledge-Intensive Tasks","papers":2595,"papersByYear":[53,65,79,130,189,220,395,624,840],"rel":2.9063,"reviewed":true,"share":[0.9446,0.9095,0.8861,1.2653,1.6073,1.5849,2.0881,3.3514,3.6899],"size":4314,"slope":0.003505},{"claims":[50,88,89,140,139,161,380,874,2253],"deltaPp":4.8181,"id":20,"keywords":["agents","ai","llm","systems","benchmark"],"label":"LLM Agent Benchmarks","labelFull":"Benchmarks and Evaluation for LLM-Based Autonomous Agents","papers":2434,"papersByYear":[38,59,66,101,109,123,225,462,1251],"rel":7.1147,"reviewed":true,"share":[0.6772,0.8255,0.7403,0.9831,0.9269,0.8861,1.1894,2.4813,5.4953],"size":4174,"slope":0.004173},{"claims":[65,122,245,369,514,608,658,640,829],"deltaPp":1.6033,"id":21,"keywords":["distillation","student","teacher","knowledge","feature"],"label":"Knowledge Distillation","labelFull":"Knowledge Distillation for Efficient Model Training","papers":2281,"papersByYear":[35,59,127,192,274,331,382,374,507],"rel":2.5702,"reviewed":true,"share":[0.6238,0.8255,1.4246,1.8688,2.3301,2.3846,2.0193,2.0087,2.2271],"size":4050,"slope":0.001945},{"claims":[8,13,12,45,87,447,923,973,1311],"deltaPp":3.3015,"id":22,"keywords":["diffusion","sampling","process","latent","generation"],"label":"Efficient Diffusion Modeling","labelFull":"Efficient Diffusion Modeling: Fast Training and Sampling","papers":2294,"papersByYear":[7,10,10,31,48,258,562,588,780],"rel":26.4543,"reviewed":true,"share":[0.1248,0.1399,0.1122,0.3017,0.4082,1.8587,2.9709,3.1581,3.4263],"size":3819,"slope":0.004922},{"claims":[364,333,424,423,475,560,498,489,250],"deltaPp":-1.9057,"id":23,"keywords":["face","facial","recognition","expression","identity"],"label":"3D Face Modeling","labelFull":"Neural 3D Face Modeling for Reconstruction and Recognition","papers":1513,"papersByYear":[137,134,168,157,182,197,206,210,122],"rel":-0.7805,"reviewed":true,"share":[2.4416,1.8749,1.8845,1.5281,1.5478,1.4192,1.089,1.1279,0.5359],"size":3816,"slope":-0.001927},{"claims":[391,425,414,459,559,399,520,359,235],"deltaPp":-2.8293,"id":24,"keywords":["entity","relation","extraction","named","recognition"],"label":"Joint Entity-Relation Extraction","labelFull":"Joint Entity and Relation Extraction from Text","papers":1865,"papersByYear":[194,207,203,213,269,200,255,181,143],"rel":-0.8183,"reviewed":true,"share":[3.4575,2.8963,2.2771,2.0732,2.2876,1.4408,1.348,0.9721,0.6282],"size":3761,"slope":-0.003263},{"claims":[116,148,267,336,356,569,684,514,634],"deltaPp":0.4286,"id":25,"keywords":["causal","counterfactual","treatment","effect","variables"],"label":"Causal Inference Benchmarking","labelFull":"Benchmarking and Evaluation of Causal Inference Methods","papers":1721,"papersByYear":[60,69,113,161,172,270,306,229,341],"rel":0.4008,"reviewed":true,"share":[1.0693,0.9654,1.2675,1.5671,1.4627,1.9451,1.6176,1.2299,1.4979],"size":3624,"slope":0.000598},{"claims":[182,215,316,366,387,514,628,443,519],"deltaPp":-0.2048,"id":26,"keywords":["super-resolution","restoration","denoising","image","deblurring"],"label":"Image Restoration","labelFull":"Deep Learning for Image Restoration Tasks","papers":2074,"papersByYear":[98,118,163,196,212,293,354,289,351],"rel":-0.1173,"reviewed":true,"share":[1.7466,1.651,1.8284,1.9077,1.8029,2.1108,1.8713,1.5522,1.5418],"size":3570,"slope":-0.000138},{"claims":[191,278,363,443,529,491,448,288,285],"deltaPp":-0.9037,"id":27,"keywords":["object","detection","detectors","coco","localization"],"label":"Object Detection","labelFull":"Object Detection Benchmarks and Methods","papers":1929,"papersByYear":[100,153,195,234,294,292,263,198,200],"rel":-0.5071,"reviewed":true,"share":[1.7822,2.1408,2.1873,2.2776,2.5002,2.1036,1.3903,1.0634,0.8785],"size":3316,"slope":-0.001436},{"claims":[13,25,69,95,176,249,440,724,1359],"deltaPp":4.091,"id":28,"keywords":["token","attention","prediction","information"],"label":"Token Pruning","labelFull":"Adaptive Token Reduction for Efficient Transformers","papers":2293,"papersByYear":[12,21,52,78,129,193,315,513,980],"rel":19.1258,"reviewed":true,"share":[0.2139,0.2938,0.5833,0.7592,1.097,1.3904,1.6652,2.7553,4.3049],"size":3150,"slope":0.004424},{"claims":[115,177,196,288,289,276,459,521,776],"deltaPp":0.6952,"id":29,"keywords":["attention","heads","mechanism","transformer","linear"],"label":"Efficient Attention","labelFull":"Efficient Attention Mechanisms for Transformers","papers":2042,"papersByYear":[83,114,132,179,198,190,307,344,495],"rel":0.47,"reviewed":true,"share":[1.4792,1.5951,1.4807,1.7423,1.6838,1.3688,1.6229,1.8476,2.1744],"size":3097,"slope":0.000575},{"claims":[96,153,194,366,471,553,485,382,315],"deltaPp":-0.4128,"id":30,"keywords":["vision","transformers","image","visual","classification"],"label":"Vision Transformers","labelFull":"Vision Transformers: Efficient and Self‑Supervised Methods","papers":2340,"papersByYear":[87,131,166,272,350,402,376,297,259],"rel":-0.2662,"reviewed":true,"share":[1.5505,1.8329,1.862,2.6475,2.9764,2.896,1.9876,1.5951,1.1377],"size":3015,"slope":-0.000311},{"claims":[419,452,422,371,346,244,347,269,143],"deltaPp":-3.727,"id":31,"keywords":["word","embeddings","text classification","sentence","sense"],"label":"Text Embeddings","labelFull":"Text Embedding Learning, Evaluation, and Analysis","papers":1753,"papersByYear":[235,259,243,210,193,146,206,156,105],"rel":-0.8899,"reviewed":true,"share":[4.1882,3.6239,2.7257,2.044,1.6413,1.0518,1.089,0.8379,0.4612],"size":3013,"slope":-0.004589},{"claims":[108,201,310,367,453,501,485,237,321],"deltaPp":-0.2911,"id":32,"keywords":["regret","games","bound","equilibrium","algorithm"],"label":"No-Regret Learning","labelFull":"No-Regret Learning for Online Algorithms","papers":1551,"papersByYear":[57,102,163,198,224,258,255,129,165],"rel":-0.2865,"reviewed":true,"share":[1.0159,1.4272,1.8284,1.9272,1.9049,1.8587,1.348,0.6928,0.7248],"size":2983,"slope":-0.000733},{"claims":[248,268,285,355,476,329,467,330,212],"deltaPp":-1.2333,"id":33,"keywords":["summarization","abstractive","extractive","document","evaluation"],"label":"Abstractive Summarization","labelFull":"Abstractive Summarization Models, Benchmarks, and Evaluation","papers":1326,"papersByYear":[101,112,118,153,188,141,215,169,129],"rel":-0.6852,"reviewed":true,"share":[1.8,1.5671,1.3236,1.4892,1.5988,1.0158,1.1365,0.9077,0.5667],"size":2970,"slope":-0.001293},{"claims":[114,179,256,303,303,383,435,358,634],"deltaPp":0.5411,"id":34,"keywords":["policy","reward","markov decision","decision processes","mdps"],"label":"Off-Policy Policy Optimization","labelFull":"Off-Policy Policy Optimization in Reinforcement Learning","papers":1967,"papersByYear":[84,118,168,204,196,228,265,240,464],"rel":0.3614,"reviewed":true,"share":[1.4971,1.651,1.8845,1.9856,1.6668,1.6425,1.4009,1.289,2.0382],"size":2965,"slope":-3.9e-05},{"claims":[288,363,378,372,339,382,335,243,260],"deltaPp":-2.2398,"id":35,"keywords":["variational","posterior","gaussian","bayesian","inference"],"label":"Variational Bayesian Deep Learning","labelFull":"Variational Inference for Bayesian Neural Nets and GPs","papers":1892,"papersByYear":[173,221,227,233,214,246,220,166,192],"rel":-0.7265,"reviewed":true,"share":[3.0832,3.0922,2.5463,2.2679,1.8199,1.7722,1.163,0.8916,0.8434],"size":2960,"slope":-0.003137},{"claims":[101,215,249,334,420,459,448,382,315],"deltaPp":0.0415,"id":36,"keywords":["few-shot","zero-shot","classification","learning","generalization"],"label":"Few-Shot & Zero-Shot Learning","labelFull":"Few-Shot and Zero-Shot Learning for Visual Classification","papers":2005,"papersByYear":[62,117,146,208,272,325,326,288,261],"rel":0.0376,"reviewed":true,"share":[1.105,1.6371,1.6377,2.0245,2.3131,2.3413,1.7233,1.5468,1.1465],"size":2923,"slope":6.4e-05},{"claims":[33,88,176,198,260,336,439,468,718],"deltaPp":1.5241,"id":37,"keywords":["flow","equations","pde","differential","neural"],"label":"Neural PDE Solvers","labelFull":"Neural Solvers for Partial Differential Equations","papers":1509,"papersByYear":[18,57,101,119,146,176,233,239,420],"rel":4.7509,"reviewed":true,"share":[0.3208,0.7975,1.1329,1.1583,1.2416,1.2679,1.2317,1.2836,1.8449],"size":2716,"slope":0.00131},{"claims":[242,272,300,293,359,331,298,362,237],"deltaPp":-1.3052,"id":38,"keywords":["depth","stereo","estimation","monocular","disparity"],"label":"Monocular Depth Estimation","labelFull":"Monocular Depth Estimation from Single Images","papers":1272,"papersByYear":[107,118,141,119,159,165,154,172,137],"rel":-0.6844,"reviewed":true,"share":[1.907,1.651,1.5816,1.1583,1.3522,1.1887,0.8141,0.9238,0.6018],"size":2694,"slope":-0.001484},{"claims":[62,105,171,208,239,300,367,476,732],"deltaPp":1.2223,"id":39,"keywords":["navigation","manipulation","robotic","locomotion","environments"],"label":"Embodied Robot Learning","labelFull":"Learning for Robotic Navigation, Manipulation, Locomotion","papers":1681,"papersByYear":[48,75,101,137,148,188,239,272,473],"rel":1.4288,"reviewed":true,"share":[0.8555,1.0494,1.1329,1.3335,1.2586,1.3544,1.2634,1.4609,2.0778],"size":2660,"slope":0.001068},{"claims":[0,4,0,15,48,355,877,767,570],"deltaPp":1.9679,"id":40,"keywords":["diffusion","text-to-image","denoising","video","latent"],"label":"Controllable Text-to-Image Diffusion","labelFull":"Controllable Text-to-Image Diffusion: Efficient Sampling","papers":1929,"papersByYear":[0,3,0,10,32,241,619,576,448],"rel":null,"reviewed":true,"share":[0.0,0.042,0.0,0.0973,0.2721,1.7362,3.2722,3.0936,1.9679],"size":2636,"slope":0.004202},{"claims":[81,144,182,246,303,420,415,323,398],"deltaPp":0.1972,"id":41,"keywords":["privacy","private","differential","membership","guarantees"],"label":"Differentially Private Machine Learning","labelFull":"Differential Privacy for Machine Learning","papers":1152,"papersByYear":[37,59,70,105,128,192,200,166,195],"rel":0.2991,"reviewed":true,"share":[0.6594,0.8255,0.7852,1.022,1.0885,1.3832,1.0573,0.8916,0.8566],"size":2512,"slope":0.000315},{"claims":[47,62,63,113,93,174,446,525,925],"deltaPp":2.3253,"id":42,"keywords":["code","instruction","program","verification","generation"],"label":"LLM Code Generation","labelFull":"LLM Code Generation, Verification, and Evaluation","papers":1588,"papersByYear":[29,35,37,57,56,101,272,354,647],"rel":4.4994,"reviewed":true,"share":[0.5168,0.4897,0.415,0.5548,0.4762,0.7276,1.4379,1.9013,2.8421],"size":2448,"slope":0.002626},{"claims":[204,229,257,303,319,278,307,226,193],"deltaPp":-1.2646,"id":43,"keywords":["action","recognition","temporal","video","activity"],"label":"Video Action Recognition","labelFull":"Video Action Recognition and Temporal Action Detection","papers":1250,"papersByYear":[103,118,125,158,159,157,160,140,130],"rel":-0.6889,"reviewed":true,"share":[1.8357,1.651,1.4021,1.5379,1.3522,1.131,0.8458,0.7519,0.5711],"size":2316,"slope":-0.001546},{"claims":[111,109,110,118,135,171,337,460,742],"deltaPp":0.7797,"id":44,"keywords":["clinical","medical","citation","review","scientific"],"label":"Biomedical LLMs","labelFull":"Biomedical Large Language Models and Evaluation Benchmarks","papers":1292,"papersByYear":[61,62,62,75,82,104,177,244,425],"rel":0.7172,"reviewed":true,"share":[1.0872,0.8675,0.6955,0.73,0.6973,0.7492,0.9357,1.3105,1.8669],"size":2293,"slope":0.000825},{"claims":[52,72,88,127,155,247,469,389,683],"deltaPp":0.8132,"id":45,"keywords":["time series","forecasting","multivariate","weather","temporal"],"label":"Time Series Foundation Models","labelFull":"Foundation Models for Time Series Forecasting","papers":1110,"papersByYear":[33,41,54,72,80,120,204,187,319],"rel":1.3828,"reviewed":true,"share":[0.5881,0.5737,0.6057,0.7008,0.6803,0.8645,1.0784,1.0044,1.4013],"size":2282,"slope":0.000942},{"claims":[13,19,16,28,166,286,501,564,507],"deltaPp":1.4905,"id":46,"keywords":["prompt","tuning","responses","learning"],"label":"Prompt Optimization","labelFull":"Prompt Optimization for Large Language Models","papers":1511,"papersByYear":[12,17,12,19,100,186,371,406,388],"rel":6.9682,"reviewed":true,"share":[0.2139,0.2379,0.1346,0.1849,0.8504,1.34,1.9612,2.1806,1.7044],"size":2100,"slope":0.002766},{"claims":[76,150,209,296,310,327,301,256,133],"deltaPp":-0.3266,"id":47,"keywords":["point","cloud","registration","3d","completion"],"label":"Point Cloud Deep Learning","labelFull":"Deep Learning for 3D Point Cloud Analysis","papers":1189,"papersByYear":[41,78,119,153,175,206,178,147,92],"rel":-0.447,"reviewed":true,"share":[0.7307,1.0914,1.3348,1.4892,1.4882,1.484,0.941,0.7895,0.4041],"size":2058,"slope":-0.000501},{"claims":[93,116,198,196,258,313,388,301,194],"deltaPp":-0.343,"id":48,"keywords":["shape","surface","3d","signed","reconstruction"],"label":"Neural Implicit 3D Reconstruction","labelFull":"Neural Implicit Representations for 3D Shape Reconstruction","papers":1326,"papersByYear":[54,70,130,120,152,206,245,208,141],"rel":-0.3564,"reviewed":true,"share":[0.9624,0.9794,1.4582,1.168,1.2926,1.484,1.2951,1.1171,0.6194],"size":2057,"slope":-0.000162},{"claims":[355,343,312,253,249,139,207,103,57],"deltaPp":-2.9073,"id":49,"keywords":["parsing","dependency","parser","morphological","constituency"],"label":"Neural Parsing","labelFull":"Neural Dependency, Constituency, and Semantic Parsing","papers":1009,"papersByYear":[172,152,150,129,126,72,104,68,36],"rel":-0.9484,"reviewed":true,"share":[3.0654,2.1268,1.6826,1.2556,1.0715,0.5187,0.5498,0.3652,0.1581],"size":2018,"slope":-0.003319},{"claims":[36,64,110,149,193,313,389,379,328],"deltaPp":0.39,"id":50,"keywords":["anomaly","detection","out-of-distribution","ood","outlier"],"label":"Anomaly Detection","labelFull":"Anomaly Detection and Out-of-Distribution Detection","papers":952,"papersByYear":[21,34,58,81,83,153,176,172,174],"rel":1.0419,"reviewed":true,"share":[0.3743,0.4757,0.6506,0.7884,0.7058,1.1022,0.9304,0.9238,0.7643],"size":1961,"slope":0.00063},{"claims":[81,117,157,133,184,212,265,346,457],"deltaPp":0.4664,"id":51,"keywords":["quantization","quantized","post-training","weight","error"],"label":"Neural Network Quantization","labelFull":"Neural Network Quantization: Post-Training and Low-Bit","papers":950,"papersByYear":[32,47,67,58,86,104,141,179,236],"rel":0.8178,"reviewed":true,"share":[0.5703,0.6576,0.7515,0.5645,0.7314,0.7492,0.7454,0.9614,1.0367],"size":1952,"slope":0.000492},{"claims":[28,50,105,155,152,182,390,334,527],"deltaPp":1.2098,"id":52,"keywords":["decoding","transformers","speculative","architecture","inference"],"label":"Efficient Transformers","labelFull":"Efficient Training and Inference for Transformer Models","papers":1429,"papersByYear":[28,45,72,116,122,136,280,241,389],"rel":2.4244,"reviewed":true,"share":[0.499,0.6296,0.8076,1.1291,1.0375,0.9798,1.4802,1.2944,1.7088],"size":1923,"slope":0.001338},{"claims":[102,162,222,282,295,299,248,175,130],"deltaPp":-0.5725,"id":53,"keywords":["domain","adaptation","generalization","source","target"],"label":"Domain Adaptation","labelFull":"Unsupervised, Semi-supervised and Open-Set Domain Adaptation","papers":1264,"papersByYear":[58,96,134,165,196,200,184,126,105],"rel":-0.5538,"reviewed":true,"share":[1.0337,1.3432,1.5031,1.606,1.6668,1.4408,0.9727,0.6767,0.4612],"size":1915,"slope":-0.000919},{"claims":[74,139,156,151,186,164,330,360,320],"deltaPp":0.1142,"id":54,"keywords":["driving","autonomous","traffic","lane","vehicle"],"label":"End-to-End Autonomous Driving","labelFull":"End-to-End Autonomous Driving: Datasets, Simulators, Evaluation","papers":943,"papersByYear":[35,57,81,78,92,91,165,176,168],"rel":0.1831,"reviewed":true,"share":[0.6238,0.7975,0.9086,0.7592,0.7824,0.6556,0.8722,0.9453,0.738],"size":1880,"slope":0.000121},{"claims":[0,8,55,128,263,395,439,272,218],"deltaPp":0.5798,"id":55,"keywords":["federated","learning","clients","personalized","communication"],"label":"Federated Learning","labelFull":"Federated Learning: Heterogeneous Data & Communication","papers":878,"papersByYear":[0,3,24,67,118,195,207,132,132],"rel":null,"reviewed":true,"share":[0.0,0.042,0.2692,0.6521,1.0035,1.4048,1.0943,0.709,0.5798],"size":1778,"slope":0.001121},{"claims":[97,118,154,248,257,214,200,236,232],"deltaPp":-0.3913,"id":56,"keywords":["video","frames","temporal","features","prediction"],"label":"Efficient Video Representation Learning","labelFull":"Efficient Video Representation Learning and Prediction","papers":1301,"papersByYear":[71,84,115,173,169,161,156,173,199],"rel":-0.3092,"reviewed":true,"share":[1.2654,1.1753,1.29,1.6839,1.4372,1.1599,0.8247,0.9292,0.8741],"size":1756,"slope":-0.000626},{"claims":[98,100,96,138,120,162,255,359,372],"deltaPp":0.311,"id":57,"keywords":["visual question","answering","video question","pairs","captioning"],"label":"Knowledge-Based VQA","labelFull":"Knowledge-Based Visual Question Answering","papers":1125,"papersByYear":[56,52,52,75,73,96,160,263,298],"rel":0.3116,"reviewed":true,"share":[0.998,0.7276,0.5833,0.73,0.6208,0.6916,0.8458,1.4125,1.309],"size":1700,"slope":0.000631},{"claims":[76,148,183,236,247,288,247,152,113],"deltaPp":-0.4005,"id":58,"keywords":["bandits","regret","multi-armed","contextual","linear"],"label":"Bandit Regret Minimization","labelFull":"Regret Minimization for Multi-Armed and Contextual Bandits","papers":861,"papersByYear":[38,74,91,119,127,155,121,73,63],"rel":-0.5914,"reviewed":true,"share":[0.6772,1.0354,1.0208,1.1583,1.08,1.1166,0.6396,0.3921,0.2767],"size":1690,"slope":-0.000723},{"claims":[74,116,132,159,217,187,238,275,260],"deltaPp":0.1044,"id":59,"keywords":["compression","image","rate-distortion","coding","video"],"label":"Learned Image Compression","labelFull":"Learned Image Compression: Rate-Distortion and Runtime","papers":936,"papersByYear":[39,60,71,85,109,102,136,152,182],"rel":0.1502,"reviewed":true,"share":[0.6951,0.8395,0.7964,0.8273,0.9269,0.7348,0.7189,0.8164,0.7995],"size":1658,"slope":1.7e-05},{"claims":[144,163,155,167,190,159,275,253,109],"deltaPp":-0.8377,"id":60,"keywords":["stance","news","detection","fake","political"],"label":"Political Stance Detection","labelFull":"Political Stance Detection in Social Media and News","papers":754,"papersByYear":[65,73,69,74,78,66,135,121,73],"rel":-0.7232,"reviewed":true,"share":[1.1584,1.0214,0.774,0.7203,0.6633,0.4755,0.7136,0.6499,0.3207],"size":1615,"slope":-0.000805},{"claims":[210,192,218,205,191,110,203,163,82],"deltaPp":-1.5665,"id":61,"keywords":["sentiment","aspect","analysis","opinion","classification"],"label":"Aspect-Based Sentiment Analysis","labelFull":"Aspect-Based Sentiment Analysis and Opinion Mining","papers":722,"papersByYear":[98,99,93,94,79,55,89,74,41],"rel":-0.8969,"reviewed":true,"share":[1.7466,1.3852,1.0432,0.9149,0.6718,0.3962,0.4705,0.3974,0.1801],"size":1574,"slope":-0.001816},{"claims":[141,184,148,145,136,169,242,227,173],"deltaPp":-0.7367,"id":62,"keywords":["captioning","image","video","dense","visual"],"label":"Image & Video Captioning","labelFull":"Image & Video Captioning: Generation, Diversity, Evaluation","papers":773,"papersByYear":[65,79,60,69,64,78,134,128,96],"rel":-0.636,"reviewed":true,"share":[1.1584,1.1054,0.673,0.6716,0.5443,0.5619,0.7084,0.6875,0.4217],"size":1565,"slope":-0.000707},{"claims":[93,144,130,220,192,186,185,189,177],"deltaPp":-0.4612,"id":63,"keywords":["style","stylization","transfer","content","image"],"label":"Arbitrary Image Style Transfer","labelFull":"Arbitrary Image Style Transfer and Stylization","papers":855,"papersByYear":[52,74,77,122,111,102,113,98,106],"rel":-0.4976,"reviewed":true,"share":[0.9268,1.0354,0.8637,1.1875,0.944,0.7348,0.5973,0.5263,0.4656],"size":1516,"slope":-0.000726},{"claims":[23,28,77,96,165,201,224,302,366],"deltaPp":0.7694,"id":64,"keywords":["continual learning","forgetting","unlearning","catastrophic","memorization"],"label":"Catastrophic Forgetting Mitigation","labelFull":"Mitigating Catastrophic Forgetting in Continual Learning","papers":930,"papersByYear":[15,20,53,57,93,125,142,189,236],"rel":2.8784,"reviewed":true,"share":[0.2673,0.2798,0.5945,0.5548,0.7909,0.9005,0.7506,1.0151,1.0367],"size":1482,"slope":0.00099},{"claims":[48,92,128,142,155,160,293,255,196],"deltaPp":-0.0003,"id":65,"keywords":["text","generation","natural","detection","evaluation"],"label":"Neural Text Generation","labelFull":"Neural Text Generation: Methods, Benchmarks, and Evaluation","papers":1000,"papersByYear":[36,66,80,91,100,113,196,172,146],"rel":-0.0005,"reviewed":true,"share":[0.6416,0.9235,0.8974,0.8857,0.8504,0.8141,1.0361,0.9238,0.6413],"size":1469,"slope":3.4e-05},{"claims":[118,155,129,141,97,156,204,190,237],"deltaPp":-0.3988,"id":66,"keywords":["low-rank","matrix","matrices","factorization","sensing"],"label":"Low-Rank Adaptation","labelFull":"Low-Rank Adaptation for Efficient Model Fine-Tuning","papers":891,"papersByYear":[66,78,82,73,68,87,126,134,177],"rel":-0.339,"reviewed":true,"share":[1.1763,1.0914,0.9198,0.7105,0.5783,0.6268,0.6661,0.7197,0.7775],"size":1427,"slope":-0.00055},{"claims":[125,110,136,170,171,172,169,175,153],"deltaPp":-0.6691,"id":67,"keywords":["tracking","trackers","multi-object","visual","state-of-the-art"],"label":"Visual Object Tracking","labelFull":"Visual Object Tracking and Multi-Object Tracking","papers":644,"papersByYear":[58,44,65,70,73,82,82,87,83],"rel":-0.6473,"reviewed":true,"share":[1.0337,0.6156,0.7291,0.6813,0.6208,0.5907,0.4335,0.4673,0.3646],"size":1381,"slope":-0.000634},{"claims":[38,38,45,52,60,80,185,335,537],"deltaPp":1.1335,"id":68,"keywords":["human","judgments","evaluation","judge","llm"],"label":"LLM-as-a-Judge","labelFull":"LLM-as-a-Judge Evaluation and Human Judgment Alignment","papers":1093,"papersByYear":[34,35,41,48,55,67,160,257,396],"rel":1.8705,"reviewed":true,"share":[0.606,0.4897,0.4599,0.4672,0.4677,0.4827,0.8458,1.3803,1.7395],"size":1370,"slope":0.001332},{"claims":[70,92,112,139,137,186,198,212,221],"deltaPp":-0.1851,"id":69,"keywords":["generative","latent","modeling","deep","distribution"],"label":"Plug-and-Play Generative Modeling","labelFull":"Plug-and-Play Generative Modeling: Model-Agnostic Adaptation","papers":1049,"papersByYear":[55,65,81,113,105,137,152,160,181],"rel":-0.1888,"reviewed":true,"share":[0.9802,0.9095,0.9086,1.0999,0.8929,0.987,0.8035,0.8593,0.7951],"size":1367,"slope":-0.000202},{"claims":[4,2,6,17,20,49,260,367,591],"deltaPp":1.4008,"id":70,"keywords":["hallucination","vision-language","reduces","mitigates","detection"],"label":"Hallucination Detection & Mitigation","labelFull":"Hallucination Detection and Mitigation in LLMs and VLMs","papers":711,"papersByYear":[2,2,6,9,15,26,135,189,327],"rel":39.3483,"reviewed":true,"share":[0.0356,0.028,0.0673,0.0876,0.1276,0.1873,0.7136,1.0151,1.4364],"size":1316,"slope":0.001659},{"claims":[91,126,146,183,180,169,199,147,71],"deltaPp":-0.6403,"id":71,"keywords":["event","extraction","argument","biomedical","relation"],"label":"Event Extraction","labelFull":"Event Extraction and Event Relation Modeling","papers":654,"papersByYear":[48,62,74,86,83,78,99,75,49],"rel":-0.7485,"reviewed":true,"share":[0.8555,0.8675,0.8301,0.8371,0.7058,0.5619,0.5233,0.4028,0.2152],"size":1312,"slope":-0.000807},{"claims":[9,10,11,17,54,124,346,350,365],"deltaPp":1.0476,"id":72,"keywords":["editing","text-to-image","generation","text-to-video","prompts"],"label":"Text-to-Image Generation & Editing","labelFull":"Unified Text-to-Image Generation and Editing via Prompt","papers":954,"papersByYear":[9,8,11,14,32,89,249,267,275],"rel":6.5312,"reviewed":true,"share":[0.1604,0.1119,0.1234,0.1363,0.2721,0.6412,1.3163,1.434,1.208],"size":1286,"slope":0.001841},{"claims":[96,85,136,108,112,104,216,181,247],"deltaPp":-0.2687,"id":73,"keywords":["emotion","recognition","affective","multimodal","classification"],"label":"Emotion Recognition","labelFull":"Emotion Recognition in Textual and Multimodal Contexts","papers":525,"papersByYear":[38,38,50,49,49,43,82,83,93],"rel":-0.3968,"reviewed":true,"share":[0.6772,0.5317,0.5609,0.4769,0.4167,0.3098,0.4335,0.4458,0.4085],"size":1285,"slope":-0.000292},{"claims":[16,50,105,130,194,264,252,155,98],"deltaPp":0.0588,"id":74,"keywords":["3d object","lidar","object detection","monocular 3d","kitti"],"label":"LiDAR 3D Object Detection","labelFull":"LiDAR-Based 3D Object Detection for Autonomous Driving","papers":708,"papersByYear":[11,30,60,69,105,150,137,88,58],"rel":0.3,"reviewed":true,"share":[0.196,0.4198,0.673,0.6716,0.8929,1.0806,0.7242,0.4726,0.2548],"size":1264,"slope":0.000151},{"claims":[9,17,13,22,28,72,260,301,499],"deltaPp":1.6406,"id":75,"keywords":["reinforcement","agents","preference","planning","learning"],"label":"LLM Reinforcement Learning","labelFull":"Reinforcement Learning for LLM Agent Alignment and Planning","papers":954,"papersByYear":[9,10,11,15,20,53,192,234,410],"rel":10.2282,"reviewed":true,"share":[0.1604,0.1399,0.1234,0.146,0.1701,0.3818,1.015,1.2568,1.801],"size":1221,"slope":0.001989},{"claims":[67,87,129,106,164,130,213,173,145],"deltaPp":-0.3491,"id":76,"keywords":["knowledge graph","graph completion","temporal knowledge","embedding","link prediction"],"label":"Knowledge Graph Completion","labelFull":"Knowledge Graph Completion and Link Prediction","papers":732,"papersByYear":[43,56,75,71,89,71,122,110,95],"rel":-0.4555,"reviewed":true,"share":[0.7664,0.7835,0.8413,0.6911,0.7569,0.5115,0.6449,0.5908,0.4173],"size":1214,"slope":-0.000424},{"claims":[43,58,62,116,106,162,171,226,253],"deltaPp":0.1008,"id":77,"keywords":["retrieval","image","grounding","video","composed"],"label":"Cross-Modal Retrieval","labelFull":"Cross-Modal Retrieval for Text-Image and Text-Video Tasks","papers":838,"papersByYear":[36,38,46,81,74,118,127,149,169],"rel":0.1571,"reviewed":true,"share":[0.6416,0.5317,0.516,0.7884,0.6293,0.8501,0.6714,0.8003,0.7424],"size":1197,"slope":0.000264},{"claims":[34,57,80,95,127,186,249,206,116],"deltaPp":-0.0085,"id":78,"keywords":["avatars","garment","clothing","hair","body"],"label":"Clothed Human Avatar Reconstruction","labelFull":"Neural Reconstruction of 3D Human Avatars with Clothing","papers":538,"papersByYear":[16,27,30,42,60,83,118,99,63],"rel":-0.0298,"reviewed":true,"share":[0.2852,0.3778,0.3365,0.4088,0.5102,0.5979,0.6238,0.5317,0.2767],"size":1150,"slope":0.000199},{"claims":[39,42,84,120,157,171,179,164,187],"deltaPp":0.0906,"id":79,"keywords":["explanations","rationales","faithfulness","explainability","human"],"label":"Faithful Explanations","labelFull":"Faithful and Human-Aligned Model Explanation Methods","papers":692,"papersByYear":[24,26,50,70,87,102,107,108,118],"rel":0.2118,"reviewed":true,"share":[0.4277,0.3638,0.5609,0.6813,0.7399,0.7348,0.5656,0.5801,0.5183],"size":1143,"slope":0.000179},{"claims":[33,22,73,88,134,143,250,137,247],"deltaPp":0.1672,"id":80,"keywords":["spiking","neural networks","neuron","plasticity","temporal"],"label":"Efficient SNN Training","labelFull":"Efficient Training Methods for Spiking Neural Networks","papers":472,"papersByYear":[16,14,27,41,54,62,100,55,103],"rel":0.5863,"reviewed":true,"share":[0.2852,0.1959,0.3029,0.3991,0.4592,0.4467,0.5286,0.2954,0.4524],"size":1127,"slope":0.000244},{"claims":[17,45,72,104,116,186,208,126,245],"deltaPp":0.6464,"id":81,"keywords":["uncertainty","calibration","out-of-distribution","epistemic","shift"],"label":"Uncertainty Quantification","labelFull":"Uncertainty Quantification and Calibration in Deep Learning","papers":894,"papersByYear":[15,28,55,78,96,150,166,98,208],"rel":2.4183,"reviewed":true,"share":[0.2673,0.3918,0.6169,0.7592,0.8164,1.0806,0.8775,0.5263,0.9137],"size":1119,"slope":0.000639},{"claims":[65,70,92,86,128,194,183,166,112],"deltaPp":-0.1148,"id":82,"keywords":["hand","gesture","pose","3d","contact"],"label":"Hand-Object Pose Estimation","labelFull":"3D Hand-Object Pose Estimation From RGB/RGB-D Images","papers":428,"papersByYear":[20,24,32,34,46,70,80,67,55],"rel":-0.3221,"reviewed":true,"share":[0.3564,0.3358,0.3589,0.3309,0.3912,0.5043,0.4229,0.3598,0.2416],"size":1096,"slope":-1.4e-05},{"claims":[84,101,111,133,128,181,126,118,108],"deltaPp":-0.4136,"id":83,"keywords":["clustering","k-means","algorithm","approximation","fair"],"label":"Scalable Clustering","labelFull":"Scalable and Efficient Clustering Algorithms","papers":534,"papersByYear":[38,49,48,70,62,82,61,64,60],"rel":-0.6108,"reviewed":true,"share":[0.6772,0.6856,0.5384,0.6813,0.5273,0.5907,0.3225,0.3437,0.2636],"size":1090,"slope":-0.000534},{"claims":[64,110,159,186,163,163,120,65,50],"deltaPp":-0.7472,"id":84,"keywords":["imagenet","top-1","accuracy","resnet-50","higher"],"label":"Efficient ImageNet Classification","labelFull":"Efficient ImageNet Classification Models","papers":934,"papersByYear":[54,91,133,156,141,142,106,62,49],"rel":-0.7764,"reviewed":true,"share":[0.9624,1.2733,1.4919,1.5184,1.1991,1.023,0.5603,0.333,0.2152],"size":1080,"slope":-0.001361},{"claims":[122,135,151,153,106,73,94,123,103],"deltaPp":-0.6944,"id":85,"keywords":["person","re-identification","identity","visible-infrared","search"],"label":"Person Re-Identification","labelFull":"Person Re-Identification Across Cameras and Modalities","papers":527,"papersByYear":[54,66,69,76,48,38,54,61,61],"rel":-0.7215,"reviewed":true,"share":[0.9624,0.9235,0.774,0.7397,0.4082,0.2738,0.2855,0.3276,0.268],"size":1060,"slope":-0.001001},{"claims":[175,207,193,196,135,80,42,12,10],"deltaPp":-1.4219,"id":86,"keywords":["gan","discriminator","generator","conditional","latent"],"label":"Stable GAN Training","labelFull":"Stable and Data-Efficient GAN Training","papers":561,"papersByYear":[82,106,98,104,82,48,22,10,9],"rel":-0.973,"reviewed":true,"share":[1.4614,1.4831,1.0993,1.0123,0.6973,0.3458,0.1163,0.0537,0.0395],"size":1050,"slope":-0.002101},{"claims":[73,114,141,156,158,163,121,61,55],"deltaPp":-0.899,"id":87,"keywords":["cifar-10","cifar-100","imagenet","mnist","cifar10"],"label":"Standard Image Classification Benchmarks","labelFull":"Benchmark Results on Standard Image Classification Datasets","papers":977,"papersByYear":[64,106,135,142,151,149,115,60,55],"rel":-0.7882,"reviewed":true,"share":[1.1406,1.4831,1.5143,1.3821,1.2841,1.0734,0.6079,0.3223,0.2416],"size":1042,"slope":-0.001533},{"claims":[34,46,48,71,90,89,176,224,252],"deltaPp":0.3779,"id":88,"keywords":["claim","legal","verification","evidence","fact"],"label":"Automated Fact-Checking","labelFull":"Automated Fact-Checking: Claim Verification and Evidence Retrieval","papers":541,"papersByYear":[17,24,29,35,40,46,86,109,155],"rel":1.2472,"reviewed":true,"share":[0.303,0.3358,0.3253,0.3407,0.3402,0.3314,0.4546,0.5854,0.6809],"size":1030,"slope":0.000418},{"claims":[2,6,12,25,37,70,267,243,343],"deltaPp":0.5137,"id":89,"keywords":["watermark","steganography","robustness","embeds","attacks"],"label":"Robust Generative AI Watermarking","labelFull":"Robust Watermarking Techniques for Generative AI Models","papers":346,"papersByYear":[1,3,6,8,14,24,84,85,121],"rel":28.8596,"reviewed":true,"share":[0.0178,0.042,0.0673,0.0779,0.1191,0.1729,0.444,0.4565,0.5315],"size":1005,"slope":0.000691},{"claims":[7,12,21,15,15,27,181,306,397],"deltaPp":1.0923,"id":90,"keywords":["preference","optimization","direct","alignment","human"],"label":"Preference Optimization","labelFull":"Preference Optimization for LLM Alignment","papers":685,"papersByYear":[6,8,18,11,12,18,113,226,273],"rel":10.218,"reviewed":true,"share":[0.1069,0.1119,0.2019,0.1071,0.102,0.1297,0.5973,1.2138,1.1992],"size":981,"slope":0.001415},{"claims":[43,50,64,79,74,103,170,190,204],"deltaPp":0.0545,"id":91,"keywords":["recommendation","user","personalization","collaborative filtering","item"],"label":"Recommender Systems","labelFull":"Recommender Systems: Algorithms, Benchmarks, LLM Approaches","papers":552,"papersByYear":[28,26,41,43,42,64,82,100,126],"rel":0.1092,"reviewed":true,"share":[0.499,0.3638,0.4599,0.4185,0.3572,0.4611,0.4335,0.5371,0.5535],"size":977,"slope":0.000121},{"claims":[69,94,104,113,133,127,99,123,69],"deltaPp":-0.5377,"id":92,"keywords":["flow","optical","scene","motion","estimation"],"label":"Optical Flow Estimation","labelFull":"Optical Flow Estimation and Scene Flow Techniques","papers":529,"papersByYear":[42,58,56,57,71,68,58,71,48],"rel":-0.7184,"reviewed":true,"share":[0.7485,0.8115,0.6282,0.5548,0.6038,0.4899,0.3066,0.3813,0.2108],"size":931,"slope":-0.000692},{"claims":[23,50,36,38,68,112,162,129,293],"deltaPp":0.3086,"id":93,"keywords":["eeg","brain","fmri","activity","recordings"],"label":"EEG Foundation Models","labelFull":"Large-Scale Pretrained EEG Foundation Models","papers":433,"papersByYear":[13,27,19,23,34,58,73,63,123],"rel":1.3319,"reviewed":true,"share":[0.2317,0.3778,0.2131,0.2239,0.2891,0.4178,0.3859,0.3384,0.5403],"size":911,"slope":0.000276},{"claims":[15,22,67,83,102,104,127,144,221],"deltaPp":0.6298,"id":94,"keywords":["trajectory","prediction","future","forecasting","state-of-the-art"],"label":"Trajectory Prediction","labelFull":"Trajectory Prediction for Autonomous Driving","papers":614,"papersByYear":[11,16,39,53,58,68,88,93,188],"rel":3.2133,"reviewed":true,"share":[0.196,0.2239,0.4375,0.5159,0.4932,0.4899,0.4652,0.4995,0.8258],"size":885,"slope":0.000563},{"claims":[44,73,86,59,83,85,169,130,149],"deltaPp":-0.1099,"id":95,"keywords":["annotation","agreement","crowdsourcing","human","quality"],"label":"Annotation Quality Control","labelFull":"Annotation Quality Control and Inter-Annotator Agreement","papers":669,"papersByYear":[35,51,64,46,62,67,121,106,117],"rel":-0.1762,"reviewed":true,"share":[0.6238,0.7136,0.7179,0.4477,0.5273,0.4827,0.6396,0.5693,0.5139],"size":878,"slope":-0.000166},{"claims":[56,35,85,51,77,68,144,139,222],"deltaPp":-0.0378,"id":96,"keywords":["story","narrative","storytelling","generation","coherence"],"label":"Story Generation","labelFull":"Neural Story Generation, Modeling, and Evaluation","papers":411,"papersByYear":[28,16,38,24,33,33,61,73,105],"rel":-0.0758,"reviewed":true,"share":[0.499,0.2239,0.4262,0.2336,0.2806,0.2377,0.3225,0.3921,0.4612],"size":877,"slope":2.5e-05},{"claims":[24,58,54,80,93,117,157,143,151],"deltaPp":0.2048,"id":97,"keywords":["pretraining","pre-training","fine-tuning","additional","reduces"],"label":"Parameter-Efficient Training","labelFull":"Parameter-Efficient Training Methods for Large-Scale Models","papers":821,"papersByYear":[24,55,51,73,90,109,142,133,144],"rel":0.4788,"reviewed":true,"share":[0.4277,0.7696,0.5721,0.7105,0.7654,0.7852,0.7506,0.7143,0.6325],"size":877,"slope":0.000181}],"totals":{"claims":[25570,33199,41765,48698,55892,66715,90760,91337,116243],"papers":[5611,7147,8915,10274,11759,13881,18917,18619,22765]},"years":[2018,2019,2020,2021,2022,2023,2024,2025,2026]}