INDEX

Explanations

keywords related to being a favorite or favorite things

oai_token-act-pair · gpt-3.5-turbo

terms related to preferences or favorites

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ural

-0.79

ijk

-0.79

idental

-0.79

absor

-0.76

inas

-0.76

abeth

-0.75

okin

-0.74

ional

-0.74

TPPStreamerBot

-0.74

asc

-0.72

POSITIVE LOGITS

 favorites

0.93

 haunt

0.85

 favorite

0.81

 Favor

0.75

é¾įå¥ĳå£«

0.74

 darling

0.73

é¾įå

0.73

 favourites

0.72

 Leave

0.70

Fav

0.70

Activations Density 0.019%

keywords related to being a favorite or favorite things

terms related to preferences or favorites

No Comments

No Known Activations