INDEX

Explanations

punctuation marks and symbols

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and symbols that denote separations or list items

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĵĺ

-0.52

NetMessage

-0.52

 cryst

-0.51

 mathemat

-0.50

 skelet

-0.48

 suspic

-0.48

 exha

-0.47

 lapt

-0.47

 antioxid

-0.47

 corrid

-0.47

POSITIVE LOGITS

ONSORED

0.46

uart

0.39

lycer

0.38

elvet

0.38

example

0.38

say

0.37

âĵĺ

0.37

 Suit

0.37

 Attend

0.36

LIST

0.36

Activations Density 0.750%

punctuation marks and symbols

punctuation marks and symbols that denote separations or list items

No Comments

No Known Activations