luisrui/KL_con0.01_PGD_alpha0.1_epsilon1e-3_niter1_224
13B
•
Updated
Modality-Interference-in-MLLMS(Trained Models)
A curated collection of trained models used in the Modality Interference in Multimodal Large Language Models (MLLMs) project.
-
-
luisrui/llava-1.5-13b-fft-KL-con0.01
13B • Updated -
luisrui/llava-1.5-13b-fft-JS-con0.01
13B • Updated -
luisrui/llava-1.5-13b-fft-JS-con1
13B • Updated -
luisrui/llava-1.5-13b-PGD_alpha0.1_epsilon1e-3_niter1_2_2_4
13B • Updated -
luisrui/llava-1.5-13b-PGD_alpha0.1_epsilon1e-3_niter2_2_2_4
13B • Updated -
luisrui/qwen7b-JS_con0.01_224_full
8B • Updated -
luisrui/KL_con0.01_PGD_alpha0.1_epsilon1e-4_niter1_224
Updated -
luisrui/qwen2.5-vl-7b-fft-full224
8B • Updated -
luisrui/qwen7b-KL_con0.01_224_full
8B • Updated -
luisrui/llava-1.5-13b-fft-2_2_4
13B • Updated -
luisrui/JS_con0.001_PGD_alpha0.1_epsilon1e-3_niter2_116
Updated -
Diagnosing and Mitigating Modality Interference in Multimodal Large Language Models
Paper • 2505.19616 • Published • 1