Tag: rl

• Paper Review: Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning (09 Jun 2025)
  
• Paper Review: Visual Planning: Lets Think Only with Images (26 May 2025)
  
• Paper Review: DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning (27 Jan 2025)
  

---

All tags

paperreview (175) deeplearning (171) nlp (70) cv (70) llm (54) transformer (35) blogpost (16) pretraining (15) sota (14) imagesegmentation (12) attention (11) pytorch (9) objectdetection (9) video (7) career (7) diffusion (6) datascience (6) agent (6) stablediffusion (5) life (5) imagegeneration (5) vlm (4) timeseries (4) selfsupervised (4) ner (4) mllm (4) languages (4) gan (4) audio (4) yolo (3) tokenization (3) superresolution (3) styletransfer (3) rnn (3) rl (3) reasoning (3) kaggle (3) imagecaptioning (3) bert (3) augmentation (3) visual (2) tts (2) transferlearning (2) simulation (2) sd (2) recommender (2) ranking (2) rag (2) qa (2) mamba (2) languagemodel (2) gpt (2) gnn (2) generation (2) fewshotlearning (2) dpo (2) distillation (2) competition (2) cnn (2) classification (2) weaksupervision (1) videogeneration (1) unet (1) textgeneration (1) tensorflow (1) tabular (1) swa (1) summarization (1) speechtranslation (1) speechtospeech (1) speechrecognition (1) speechgeneration (1) sentenceembeddings (1) semisupervised (1) scaling (1) robustness (1) robotics (1) relationextrction (1) relationextraction (1) reinforcementlearning (1) recurrent (1) recommendation (1) realtime (1) quantization (1) promptengineering (1) objecttracking (1) objectdetecion (1) nlg (1) nas (1) multimodal (1) motivation (1) motiontracking (1) mlp (1) mentoring (1) memoryoptimization (1) languagetranslation (1) jigsaw (1) instructlearning (1) inferencespeed (1) imagetextmatching (1) imagerestoration (1) imageinpainting (1) graphneuralnets (1) graph (1) forecasting (1) fail (1) evaluation (1) entitylinking (1) endtoend (1) embedding (1) efficiency (1) diffusionmodels (1) depthestimation (1) curriculumlreaning (1) contrastivelearning (1) coco (1) clip (1) chatbot (1) captioning (1) books (1) autoencoder (1) asr (1) annotation (1) anchorfree (1) alignment (1) advice (1) adversarial (1) activationfunction (1) CV (1)