INDEX

Explanations

phrases related to returning to a specific point in time or a previous state

oai_token-act-pair · gpt-3.5-turbo

references to going back to a previous state or period

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Magikarp

-0.72

SPONSORED

-0.71

erie

-0.71

reau

-0.69

erity

-0.68

ocument

-0.68

ILA

-0.67

oS

-0.67

illion

-0.67

ancial

-0.67

POSITIVE LOGITS

 basics

1.20

 fold

1.10

 roots

1.05

 drawing

1.02

 original

1.00

 originals

0.93

 forefront

0.90

 fray

0.86

 normal

0.84

 fundamentals

0.82

Activations Density 0.125%

phrases related to returning to a specific point in time or a previous state

references to going back to a previous state or period

No Comments

No Known Activations