INDEX

Explanations

information about events being publicized or made known

oai_token-act-pair · gpt-3.5-turbo

phrases related to reports and announcements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

isode

-0.77

avorite

-0.70

apeake

-0.67

venant

-0.65

appings

-0.65

Exit

-0.62

 vulner

-0.61

vette

-0.61

xtap

-0.59

umbn

-0.59

POSITIVE LOGITS

 that

0.84

that

0.72

 raining

0.68

 impossible

0.68

 unanimously

0.65

 advisable

0.61

 folly

0.59

 amongst

0.59

 fashionable

0.58

 THAT

0.56

Activations Density 0.106%

information about events being publicized or made known

phrases related to reports and announcements

No Comments

No Known Activations