INDEX

Explanations

mentions of the term "ado"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the substring "ado."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

utive

-0.77

 detrim

-0.69

Ö¼

-0.64

 Harm

-0.64

 divide

-0.64

IONS

-0.64

 mathemat

-0.60

 Hodg

-0.60

 clipboard

-0.59

ĸļ

-0.59

POSITIVE LOGITS

ado

1.20

quez

0.97

quit

0.87

osta

0.87

inho

0.79

eering

0.79

illo

0.79

ques

0.78

dar

0.74

quer

0.74

Activations Density 0.006%

mentions of the term "ado"

occurrences of the substring "ado."

No Comments

No Known Activations