INDEX

Explanations

references to direct speech involving the word "You" in different contexts

oai_token-act-pair · gpt-3.5-turbo

direct address or references to "you."

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

¿½

-0.98

£ı

-0.75

tains

-0.70

ipal

-0.69

 Commerce

-0.68

enges

-0.65

 Additional

-0.65

ŃĶ

-0.64

abin

-0.64

POSITIVE LOGITS

're

1.49

 guys

1.30

 gotta

1.26

 bastard

1.25

 deserve

1.21

've

1.17

 wanna

1.15

 idiots

1.14

owe

1.12

 idiot

1.12

Activations Density 0.130%

references to direct speech involving the word "You" in different contexts

direct address or references to "you."

No Comments

No Known Activations