INDEX

Explanations

phrases related to people's names

oai_token-act-pair · gpt-3.5-turbo

instances of the term "arg" indicating arguments or discussions

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

 Giov

-0.75

HAEL

-0.72

lihood

-0.67

 Fever

-0.67

 tremend

-0.66

utral

-0.66

 fertility

-0.65

FORMATION

-0.63

PER

-0.63

lights

-0.62

POSITIVE LOGITS

uably

1.44

uments

1.19

roup

1.03

emouth

1.02

ansas

1.00

uin

0.95

raph

0.90

uable

0.90

oths

0.89

atis

0.87

Activations Density 0.009%

phrases related to people's names

instances of the term "arg" indicating arguments or discussions

No Comments

No Known Activations