INDEX

Explanations

the word "bad" with varying levels of intensity

oai_token-act-pair · gpt-3.5-turbo

expressions related to negative evaluations or criticisms

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

ĸļ

-0.91

eters

-0.84

arov

-0.82

ittees

-0.76

ittee

-0.76

ando

-0.74

olate

-0.74

ipel

-0.74

gow

-0.73

ktop

-0.73

POSITIVE LOGITS

dest

1.00

dies

0.91

 karma

0.86

die

0.84

enough

0.82

vib

0.81

 luck

0.81

 adolesc

0.79

 manners

0.77

 enough

0.76

Activations Density 0.022%

the word "bad" with varying levels of intensity

expressions related to negative evaluations or criticisms

No Comments

No Known Activations