INDEX

Explanations

phrases indicating strong positive associations or connections

oai_token-act-pair · gpt-3.5-turbo

instances of the word "with" indicating relationships or associations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

chuk

-0.81

ãģ®éŃĶ

-0.78

tick

-0.77

arse

-0.75

earth

-0.74

hid

-0.72

BP

-0.72

ulf

-0.71

Strange

-0.70

ä¸ī

-0.70

POSITIVE LOGITS

 regard

1.14

 apologies

1.08

 regards

1.02

 plenty

0.89

draw

0.88

 exceptions

0.87

 hindsight

0.85

stood

0.85

 impunity

0.80

 instructions

0.80

Activations Density 0.105%

phrases indicating strong positive associations or connections

instances of the word "with" indicating relationships or associations

No Comments

No Known Activations