INDEX

Explanations

words related to things being wrapped around

oai_token-act-pair · gpt-3.5-turbo

words related to profoundness or depth

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

uters

-0.88

perty

-0.78

 skelet

-0.77

å¥³

-0.74

utic

-0.74

aptic

-0.73

bil

-0.70

 toget

-0.70

 exha

-0.68

EStream

-0.67

POSITIVE LOGITS

ers

0.78

er

0.74

 Rahman

0.71

Sign

0.70

ertain

0.70

rums

0.69

ation

0.68

llo

0.68

 Lore

0.67

ead

0.67

Activations Density 0.012%

words related to things being wrapped around

words related to profoundness or depth

No Comments

No Known Activations