Jonas Geiping
~Jonas_Geiping1
24
论文总数
12.0
年均投稿
平均评分
接收情况14/24
会议分布
ICLR
13
NeurIPS
5
ICML
3
COLM
3
发表论文 (24 篇)
202510 篇
4
Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach
NeurIPS 2025Spotlight
4
What do we learn from inverting CLIP models?
ICLR 2025withdrawn
4
When, Where and Why to Average Weights?
ICML 2025Poster
4
Can Language Models Falsify? Evaluating Algorithmic Reasoning with Counterexample Creation
COLM 2025Poster
3
Data Brittleness Estimation with Self-Supervised Features
ICLR 2025withdrawn
5
Gradient-based Optimization of Dataset Mixtures
ICLR 2025withdrawn
4
A Realistic Threat Model for Large Language Model Jailbreaks
ICLR 2025Rejected
4
An Interpretable N-gram Perplexity Threat Model for Large Language Model Jailbreaks
ICML 2025Poster
-
Extracting Parameter Counts from Fine-Tuning APIs
ICLR 2025withdrawn
4
Great Models Think Alike and this Undermines AI Oversight
ICML 2025Spotlight
202414 篇
3
On the Reliability of Watermarks for Large Language Models
ICLR 2024Poster
4
CALVIN: Improved Contextual Video Captioning via Instruction Tuning
NeurIPS 2024Poster
4
Privacy Backdoors: Enhancing Membership Inference through Poisoning Pre-trained Models
NeurIPS 2024Poster
4
A Simple and Efficient Baseline for Data Attribution on Images
ICLR 2024withdrawn
4
LMD3: Language Model Data Density Dependence
COLM 2024Poster
4
Universal Guidance for Diffusion Models
ICLR 2024Poster
4
Spotting LLMs With Binoculars: Zero-Shot Detection of Machine-Generated Text
ICLR 2024Rejected
4
Bring Your Own Data! Self-Sensitivity Evaluation for Large Language Models
COLM 2024Poster
4
Bring Your Own Data! Self-Supervised Evaluation for Large Language Models
ICLR 2024withdrawn
3
Transformers Can Do Arithmetic with the Right Embeddings
NeurIPS 2024Poster
4
Be like a Goldfish, Don't Memorize! Mitigating Memorization in Generative LLMs
NeurIPS 2024Poster
4
Baseline Defenses for Adversarial Attacks Against Aligned Language Models
ICLR 2024Rejected
4
Algorithm Design for Learned Algorithms
ICLR 2024withdrawn
4
NEFTune: Noisy Embeddings Improve Instruction Finetuning
ICLR 2024Poster