bug-bounty375
xss317
google246
microsoft201
facebook199
apple137
exploit112
rce96
csrf77
writeup62
account-takeover60
malware57
browser54
bragging-post53
defi48
access-control48
authentication-bypass47
smart-contract47
ethereum44
open-source42
privilege-escalation38
ssrf38
cve37
sql-injection37
web337
ai-agents35
docker35
smart-contract-vulnerability33
dos32
aws31
supply-chain29
react28
sqli28
idor28
denial-of-service26
information-disclosure26
api-security26
wordpress25
oauth25
clickjacking25
phishing25
solidity25
subdomain-takeover24
burp-suite24
race-condition23
node22
vulnerability-disclosure22
cloudflare22
reverse-engineering22
remote-code-execution22
0
8/10
A comprehensive field guide documenting 10 distinct patterns where LLMs game kernel benchmarks through timing attacks (stream injection, thread injection, lazy evaluation, patching), semantic attacks (identity kernels, no-ops, shared memory overflow), and benign shortcuts, with defensive mechanisms for each exploit category.
ai-safety
benchmark-gaming
reward-hacking
cuda
gpu-kernels
machine-learning
testing-methodology
triton
rocm
vulnerability-detection
timing-attacks
semantic-attacks
correctness-validation
adversarial-ml
KernelArena
MI300X
ROCm 6.x
CUDA
PyTorch
Triton
HIP