INDEX

Explanations

phrases related to infrequent events or occurrences

oai_token-act-pair · gpt-3.5-turbo

references to rare or unique experiences

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

ORTS

-0.67

essen

-0.65

bryce

-0.64

DD

-0.63

tions

-0.61

 Tyler

-0.60

Flow

-0.60

ãĥ¼

-0.60

RN

-0.60

Ø§ÙĦ

-0.60

POSITIVE LOGITS

 awhile

1.02

thinkable

0.70

isner

0.69

chance

0.67

theless

0.64

ocene

0.64

 Highlands

0.63

avior

0.62

 paradise

0.61

ividual

0.61

Activations Density 0.060%

phrases related to infrequent events or occurrences

references to rare or unique experiences

No Comments

No Known Activations