INDEX

Explanations

instances where things are divided or split into two or more parts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "split" and related terms indicating division or separation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

trak

-0.73

FORE

-0.68

enegger

-0.68

agara

-0.68

ties

-0.67

sburgh

-0.66

culosis

-0.64

jamin

-0.64

ECH

-0.62

 appropriated

-0.62

POSITIVE LOGITS

 hairs

1.26

 evenly

0.92

creen

0.87

 between

0.86

ters

0.79

 halves

0.79

 apart

0.74

 opinion

0.71

iple

0.71

ittal

0.71

Activations Density 0.041%

instances where things are divided or split into two or more parts

instances of the word "split" and related terms indicating division or separation

No Comments

No Known Activations