Philip Gage
The originator of the Byte Pair Encoding (BPE) algorithm, which he proposed in 1994 for data compression. The algorithm was later widely adopted in natural language processing and as a tokenization method for large language models.
Core Facts
Timeline (last 90 days)
BPE was originally a data compression algorithm proposed by Philip Gage in 1994: repeatedly find the most common adjacent byte pair in the data and replace it with a newly added unused symbol.
BPE (Byte Pair Encoding) was originally proposed by Philip Gage in 1994 for data compression, later introduced into NLP by OpenAI, and is currently the most commonly used tokenization algorithm in mainstream LLMs.
The BPE (Byte Pair Encoding) algorithm was originally proposed by Philip Gage in 1994 for data compression, and was later introduced by OpenAI into the NLP domain for tokenization in the GPT series of models.
All Facts (4)
Byte Pair Encoding (BPE) was proposed by Philip Gage in 1994 and was later introduced into the field of neural machine translation by Sennrich et al. in 2016.
70%UnverifiedBPE was originally a data compression algorithm proposed by Philip Gage in 1994: repeatedly find the most common adjacent byte pair in the data and replace it with a newly added unused symbol.
50%UnverifiedBPE (Byte Pair Encoding) was originally proposed by Philip Gage in 1994 for data compression, later introduced into NLP by OpenAI, and is currently the most commonly used tokenization algorithm in mainstream LLMs.
50%UnverifiedThe BPE (Byte Pair Encoding) algorithm was originally proposed by Philip Gage in 1994 for data compression, and was later introduced by OpenAI into the NLP domain for tokenization in the GPT series of models.
50%