GitHub - terarachang/MemPi: localize a memorized sequence in LLMs (NAACL 2024) (original) (raw)
Do Localization Methods Actually Localize Memorized Data in LLMs?
A Tale of Two Benchmarks (NAACL 2024)
Ting-Yun Chang, Jesse Thomason, and Robin Jia
🎞️ https://www.youtube.com/watch?v=V2i8CemZZHQ
Content
- Quick Start:
$ pip install -r requirements.txt - INJ Benchmark
- Data
- Information Injection
- Run Localization Methods
- DEL Benchmark
- Data
- Run Localization Methods
INJ Benchmark
Data
- Data Source : ECBD dataset from Onoe et al., 2022, see README
- Preprocessed Data: data/ecbd
Information Injection
$ bash script/ecbd/inject.sh MODEL
- MODEL: gpt2, gpt2-xl, EleutherAI/pythia-2.8b-deduped-v0, EleutherAI/pythia-6.9b-deduped
- We release our collected data at data/pile/EleutherAI
Run Localization Methods
$ bash script/ecbd/METHOD_NAME.sh MODEL
- e.g.,
bash script/ecbd/HC.sh EleutherAI/pythia-6.9b-deduped - METHOD_NAME
DEL Benchmark
Data
Find data memorized by Pythia models from the Pile-dedup
- Data Source: Please follow EleutherAI's instructions to download pretrained data in batches
- Identify memorized data with our filters:
$ bash script/pile/find.sh MODEL - We release our collected data at data/pile
Data memorized by GPT2-XL
- We release our manually collected data at data/manual/memorized_data-gpt2-xl.jsonl
Pretrained sequences for perplexity
- We randomly sample 2048 sequences from the Pile-dedupe to calculate perplexity
- shared by all LLMs
- Tokenized data at
data/pile/*/pile_random_batch.pt
Run Localization Methods
$ bash script/pile/METHOD_NAME.sh MODEL
- For Pythia models
- METHOD_NAME
$ bash script/manual/METHOD_NAME.sh
- For GPT2-XL