INDEX

Explanations

the number '2' or 'two' in various contexts

oai_token-act-pair · gpt-3.5-turbo

the word "two" and its variations, indicating references to the number two

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ugu

-0.78

 Caption

-0.75

lus

-0.71

ICLE

-0.70

lov

-0.69

asta

-0.69

amaru

-0.66

annel

-0.65

ifice

-0.65

zyme

-0.64

POSITIVE LOGITS

 thirds

1.73

 halves

1.21

 dozen

1.16

 weeks

1.09

 hundred

1.06

nd

1.06

fold

1.06

teenth

0.91

 tablespoons

0.90

thirds

0.89

Activations Density 0.128%

the number '2' or 'two' in various contexts

the word "two" and its variations, indicating references to the number two

No Comments

No Known Activations