Sotabase

Career

· Research Scientist (RS Intern), Meta2025–

Publications (17)

Why Do Multi-Agent LLM Systems Fail?
arXiv.org · 2025
200
cited
GCR: Gradient Coreset based Replay Buffer Selection for Continual Learning
Computer Vision and Pattern Recognition · 2021
168
cited
Interactive Concept Bottleneck Models
AAAI Conference on Artificial Intelligence · 2022
76
cited
ChipNet: Budget-Aware Pruning with Heaviside Continuous Approximations
International Conference on Learning Representations · 2021
33
cited
The Art of Scaling Reinforcement Learning Compute for LLMs
arXiv.org · 2025
25
cited
QuantSpec: Self-Speculative Decoding with Hierarchical Quantized KV Cache
International Conference on Machine Learning · 2025
9
cited
Using Early Readouts to Mediate Featural Bias in Distillation
IEEE Workshop/Winter Conference on Applications of Computer Vision · 2023
9
cited
Dynamic Kernel Selection for Improved Generalization and Memory Efficiency in Meta-learning
Computer Vision and Pattern Recognition · 2022
6
cited
Multi-plateau Ensemble For Endoscopic Artefact Segmentation And Detection
EndoCV@ISBI · 2020
3
cited
SPECS: Faster Test-Time Scaling through Speculative Drafts
arXiv.org · 2025
3
cited
XQuant: Breaking the Memory Wall for LLM Inference with KV Cache Rematerialization
arXiv.org · 2025
2
cited
On Designing Light-Weight Object Trackers Through Network Pruning: Use CNNS or Transformers?
IEEE International Conference on Acoustics, Speech, and Signal Processing · 2022
1
cited
A N ADVERSARIAL FEATURE LEARNING STRATEGY FOR DEBIASING NEURAL NETWORKS
2023
Arbitrage: Efficient Reasoning via Advantage-Aware Speculation
arXiv.org · 2025
RCV2023 Challenges: Benchmarking Model Training and Inference for Resource-Constrained Deep Learning
2023 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) · 2023
Rescaling CNN Through Learnable Repetition of Network Parameters
International Conference on Information Photonics · 2021
Reward Under Attack: Evaluating the Sensitivity of Process Reward Models
Sotabase
Rishabh Tiwari | Researcher Profile | Sotabase | Sotabase