List your company

Prompt Injection on GSM8k

3.79ASR@3G-Safeguard
3.16847.364211.5615.7558Aug 11, 2025Sep 2, 2025Sep 24, 2025Oct 16, 2025Nov 7, 2025Nov 29, 2025Dec 21, 2025

Evaluation results

MethodLinks
G-SafeguardTopology=Random, Backbone LLM=GPT-4o-mini2025.083.790.9911
G-SafeguardTopology=Chain, Backbone LLM=GPT-4o-mini2025.084.40.9822
G-SafeguardTopology=Tree2025.1260.9778
G-SafeguardTopology=Star, Backbone LLM=GPT-4o-mini2025.086.10.9822
BlindGuardTopology=Random, Backbone LLM=GPT-4o-mini2025.086.440.7556
G-SafeguardTopology=Star2025.126.670.9911
DOMINANTTopology=Tree, Backbone LLM=GPT-4o-mini2025.086.780.6844
BlindGuardTopology=Star, Backbone LLM=GPT-4o-mini2025.086.780.7022
G-SafeguardTopology=Tree, Backbone LLM=GPT-4o-mini2025.087.80.9911
DOMINANTTopology=Star, Backbone LLM=GPT-4o-mini2025.087.80.6978
TAMTopology=Tree, Backbone LLM=GPT-4o-mini2025.088.140.5467
TAMTopology=Star, Backbone LLM=GPT-4o-mini2025.088.140.6844
DOMINANTTopology=Chain, Backbone LLM=GPT-4o-mini2025.088.470.6756
BlindGuardTopology=Chain, Backbone LLM=GPT-4o-mini2025.088.470.6933
PREMTopology=Tree, Backbone LLM=GPT-4o-mini2025.088.470.5333
BlindGuardTopology=Tree, Backbone LLM=GPT-4o-mini2025.088.470.5955
XG-GuardTopology=Chain2025.128.670.9778
PREMTopology=Chain, Backbone LLM=GPT-4o-mini2025.088.790.6222
TAMTopology=Chain, Backbone LLM=GPT-4o-mini2025.088.840.5156
PREMTopology=Tree2025.128.890.6296
G-SafeguardTopology=Chain2025.129.330.9911
XG-GuardTopology=Tree2025.129.670.9733
TAMTopology=Star2025.129.670.7111
G-SafeguardTopology=Random2025.129.670.9911
DOMINANTTopology=Star2025.129.830.6701
No DefenseTopology=Chain, Backbone LLM=GPT-4o-mini2025.089.83—
DOMINANTTopology=Tree2025.1210.180.6897
No DefenseTopology=Tree, Backbone LLM=GPT-4o-mini2025.0810.2—
PREMTopology=Chain2025.1210.290.5962
BlindGuardTopology=Star2025.1210.330.7067
XG-GuardTopology=Random2025.1210.330.9867
DOMINANTTopology=Random2025.1210.510.6881
PREMTopology=Star, Backbone LLM=GPT-4o-mini2025.0810.510.5956
DOMINANTTopology=Random, Backbone LLM=GPT-4o-mini2025.0810.510.6978
PREMTopology=Random, Backbone LLM=GPT-4o-mini2025.0810.510.6978
BlindGuardTopology=Chain2025.1210.670.6533
DOMINANTTopology=Chain2025.1210.910.6655
TAMTopology=Chain2025.1211.670.5289
TAMTopology=Tree2025.1211.670.5467
XG-GuardTopology=Star2025.1211.670.9733
No DefenseTopology=Star, Backbone LLM=GPT-4o-mini2025.0812.89—
PREMTopology=Star2025.1213.080.641
No DefenseTopology=Chain2025.1213.67—
No DefenseTopology=Tree2025.1213.67—
BlindGuardTopology=Tree2025.12140.5599
BlindGuardTopology=Random2025.1214.330.7422
No DefenseTopology=Random, Backbone LLM=GPT-4o-mini2025.0814.48—
TAMTopology=Random, Backbone LLM=GPT-4o-mini2025.0814.910.4622
No DefenseTopology=Star2025.1215—
TAMTopology=Random2025.1215.670.4844
PREMTopology=Random2025.1216.140.6257
No DefenseTopology=Random2025.1219.33—