Research Trend Radar 100 active topics
week-over-week growth · computed from paper titles & abstracts · not citation-based
| # | Topic | This Week | Last Week | WoW |
|---|---|---|---|---|
| 1 | prediction | 52 | 64 | -18% |
| 2 | graph | 51 | 71 | -28% |
| 3 | visual | 49 | 63 | -22% |
| #4 | latent | 45 | 55 | -18% |
| #5 | memory | 45 | 70 | -35% |
| #6 | multi-agent | 43 | 56 | -23% |
| #7 | adaptive | 42 | 56 | -25% |
| #8 | inference | 41 | 42 | -2% |
| #9 | generative | 39 | 55 | -29% |
| #10 | decision | 38 | 45 | -15% |
| #11 | transformer | 37 | 41 | -10% |
| #12 | modeling | 35 | 53 | -33% |
| #13 | control | 34 | 52 | -34% |
| #14 | benchmarking | 33 | 61 | -45% |
| #15 | retrieval | 32 | 45 | -28% |
| #16 | segmentation | 31 | 46 | -32% |
| #17 | distillation | 31 | 50 | -37% |
| #18 | semantic | 30 | 49 | -38% |
| #19 | medical | 29 | 34 | -14% |
| #20 | gaussian | 28 | 36 | -22% |
| #21 | dynamics | 27 | 26 | +4% |
| #22 | foundation | 27 | 52 | -47% |
| #23 | linear | 26 | 20 | +29% |
| #24 | temporal | 26 | 29 | -10% |
| #25 | selection | 26 | 40 | -34% |
| #26 | routing | 25 | 25 | +0% |
| #27 | decoding | 25 | 35 | -28% |
| #28 | adaptation | 25 | 43 | -41% |
| #29 | robot | 24 | 21 | +14% |
| #30 | grounding | 24 | 30 | -19% |
| #31 | policy | 24 | 31 | -22% |
| #32 | estimation | 24 | 32 | -24% |
| #33 | forecasting | 24 | 37 | -34% |
| #34 | flow | 24 | 40 | -39% |
| #35 | vision-language | 24 | 48 | -49% |
| #36 | hierarchical | 23 | 22 | +4% |
| #37 | context | 23 | 32 | -27% |
| #38 | search | 23 | 36 | -35% |
| #39 | attention | 23 | 39 | -40% |
| #40 | scaling | 23 | 41 | -43% |
| #41 | feature | 23 | 42 | -44% |
| #42 | classification | 23 | 46 | -49% |
| #43 | dynamic | 22 | 30 | -26% |
| #44 | sparse | 22 | 33 | -32% |
| #45 | geometry | 21 | 22 | -4% |
| #46 | autonomous | 21 | 28 | -24% |
| #47 | matching | 21 | 32 | -33% |
| #48 | recognition | 21 | 37 | -42% |
| #49 | reconstruction | 21 | 38 | -44% |
| #50 | synthesis | 20 | 26 | -22% |
| #51 | federated | 20 | 28 | -28% |
| #52 | token | 20 | 30 | -32% |
| #53 | long-horizon | 20 | 33 | -38% |
| #54 | interaction | 20 | 36 | -43% |
| #55 | action | 20 | 37 | -45% |
| #56 | matter | 19 | 10 | +82% |
| #57 | structured | 19 | 15 | +25% |
| #58 | spatial | 19 | 22 | -13% |
| #59 | continual | 19 | 26 | -26% |
| #60 | alignment | 19 | 35 | -44% |
| #61 | improving | 18 | 7 | +138% |
| #62 | verification | 18 | 16 | +12% |
| #63 | perception | 18 | 17 | +6% |
| #64 | structure | 18 | 19 | -5% |
| #65 | skill | 18 | 20 | -10% |
| #66 | causal | 18 | 23 | -21% |
| #67 | object | 18 | 27 | -32% |
| #68 | motion | 18 | 27 | -32% |
| #69 | online | 18 | 31 | -41% |
| #70 | reward | 18 | 32 | -42% |
| #71 | recommendation | 17 | 13 | +29% |
| #72 | adversarial | 17 | 16 | +6% |
| #73 | splatting | 17 | 16 | +6% |
| #74 | tracking | 17 | 18 | -5% |
| #75 | architecture | 17 | 19 | -10% |
| #76 | scalable | 17 | 20 | -14% |
| #77 | bayesian | 17 | 20 | -14% |
| #78 | speech | 17 | 21 | -18% |
| #79 | hybrid | 17 | 22 | -22% |
| #80 | clinical | 17 | 23 | -25% |
| #81 | on-policy | 17 | 24 | -28% |
| #82 | code | 17 | 31 | -44% |
| #83 | web | 16 | 9 | +70% |
| #84 | uncertainty | 16 | 11 | +42% |
| #85 | gaussian splatting | 16 | 14 | +13% |
| #86 | streaming | 16 | 18 | -11% |
| #87 | vision | 16 | 23 | -29% |
| #88 | scientific | 16 | 24 | -32% |
| #89 | fusion | 16 | 28 | -41% |
| #90 | safety | 16 | 31 | -47% |
| #91 | active | 15 | 12 | +23% |
| #92 | self-supervised | 15 | 12 | +23% |
| #93 | selective | 15 | 14 | +7% |
| #94 | joint | 15 | 16 | -6% |
| #95 | fast | 15 | 18 | -16% |
| #96 | rethinking | 15 | 20 | -24% |
| #97 | editing | 15 | 23 | -33% |
| #98 | planning | 15 | 26 | -41% |
| #99 | discovery | 15 | 33 | -53% |
| #100 | space | 15 | 34 | -54% |