INDEX

Explanations

expressions related to methods, manners or aspects of doing something

oai_token-act-pair · gpt-3.5-turbo

phrases indicating methods or approaches

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

icio

-0.74

usters

-0.71

oute

-0.68

oglu

-0.66

ropolitan

-0.64

fif

-0.64

dinand

-0.62

inently

-0.61

 COUR

-0.61

incinn

-0.61

POSITIVE LOGITS

finding

0.85

fare

0.75

CHO

0.74

ward

0.71

station

0.71

WARD

0.70

ãĤ¨

0.69

Adapt

0.68

NE

0.68

points

0.68

Activations Density 0.033%

expressions related to methods, manners or aspects of doing something

phrases indicating methods or approaches

No Comments

No Known Activations