INDEX

Explanations

proper nouns, particularly names of people and places with the word "went."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "went."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Horus

-0.72

eers

-0.71

creen

-0.71

ament

-0.70

ribution

-0.68

uctor

-0.67

icio

-0.66

ipation

-0.64

stood

-0.63

role

-0.63

POSITIVE LOGITS

Ń·

0.97

vt

0.84

 Forth

0.81

 ahead

0.80

 overboard

0.79

lems

0.78

verning

0.78

XL

0.77

 ashore

0.77

":"/

0.76

Activations Density 0.052%

proper nouns, particularly names of people and places with the word "went."

occurrences of the word "went."

No Comments

No Known Activations