INDEX

Explanations

words related to surprise or dismay

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "shock" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

•layer 4 - surprise

No Comments

Negative Logits

amins

-0.88

uties

-0.80

allery

-0.78

umbers

-0.77

 ende

-0.71

 livest

-0.70

igion

-0.69

ittees

-0.69

guyen

-0.68

inances

-0.67

POSITIVE LOGITS

wave

1.13

waves

1.12

 absor

0.97

ingly

0.93

tro

0.88

imaru

0.87

mong

0.84

crow

0.80

Shock

0.79

shock

0.78

Activations Density 0.016%

words related to surprise or dismay

occurrences of the word "shock" and its variations

No Comments

No Known Activations