INDEX

Explanations

references to the USA or related entities

oai_token-act-pair · gpt-3.5-turbo

references to the United States or its variations in spelling

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

rained

-0.96

dress

-0.85

croft

-0.84

rums

-0.81

bread

-0.80

smanship

-0.79

osc

-0.73

sheet

-0.71

clud

-0.71

lers

-0.71

POSITIVE LOGITS

BIL

1.06

terday

1.00

WAYS

0.96

BILITY

0.96

igslist

0.92

icago

0.86

ppa

0.84

velength

0.81

UNCH

0.81

xon

0.80

Activations Density 0.015%

references to the USA or related entities

references to the United States or its variations in spelling

No Comments

No Known Activations