INDEX

Explanations

the word "so" followed by a high level of intensity or quantity descriptor

oai_token-act-pair · gpt-3.5-turbo

the word "so" in various contexts and frequencies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ICA

-0.68

 burner

-0.63

smith

-0.61

ULTS

-0.61

works

-0.59

nings

-0.56

ysc

-0.56

Nex

-0.55

amac

-0.54

eties

-0.53

POSITIVE LOGITS

oooo

1.08

othes

1.03

 much

0.99

bered

0.97

oooooooo

0.92

apy

0.91

 badly

0.91

oner

0.90

ooo

0.89

far

0.88

Activations Density 0.083%

the word "so" followed by a high level of intensity or quantity descriptor

the word "so" in various contexts and frequencies

No Comments

No Known Activations