Adobe Research
PulseAugur coverage of Adobe Research — every cluster mentioning Adobe Research across labs, papers, and developer communities, ranked by signal.
-
New method reconstructs LLM prompts from output alone
Researchers from IIT Bombay and Adobe Research have developed a method called Previous-Token Prediction (PTP) that can reconstruct a model's original prompt from its generated text alone. This technique trains an invers…
-
Researchers can reverse-engineer LLM prompts; Google reshuffles AI leadership
Researchers from IIT Bombay and Adobe Research have developed a technique called Previous-Token Prediction that can accurately reconstruct the original prompt given an LLM's output. This method functions without requiri…
-
New AI method reconstructs system instructions; Cathie Wood questions open-source risks
Researchers from the Indian Institute of Technology Bombay and Adobe Research have developed a new method called PTP that can reconstruct an AI's secret system instructions with nearly perfect accuracy by analyzing its …
-
New method reconstructs LLM prompts from output text with high accuracy
Researchers have developed a new method capable of reconstructing the original prompts used to generate text from large language models (LLMs) with high accuracy. This technique, dubbed "Previous-Token Prediction," func…
-
Sudu Technology unveils embodied AI platform with advanced skill acquisition · 1 source tracked
Sudu Technology, a year-old startup, has showcased its embodied AI platform, Sudo R1, demonstrating advanced capabilities in object manipulation and task execution. Founded by Professor Su Hao, a prominent figure in 3D …
-
Adobe, Tencent Alums Launch AI Animation Startup
A new AI animation startup has been launched by former Adobe Research scientists and Tencent T15. The company aims to provide tools for professional animation studios, with initial funding in the tens of millions from i…
-
3D Vision Research Advances Spatial Understanding and Dynamic Scene Generation
Researchers are pushing the boundaries of 3D vision, moving beyond simple reconstruction to focus on spatial understanding, dynamic simulation, and practical engineering applications. New methods are enabling models to …
-
CVPR 2026: Visual AI shifts from accuracy to understanding imperfect real-world data
Computer vision research is shifting from optimizing performance on benchmarks to enabling models to understand the world under imperfect conditions. Recent work presented around CVPR 2026 challenges fundamental assumpt…