INDEX

Explanations

titles or positions related to leadership or authority

oai_token-act-pair · gpt-3.5-turbo

phrases containing the term "head of" in various contexts related to authority or leadership positions

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

Joined

-0.69

nee

-0.68

 hereafter

-0.66

ureen

-0.66

aire

-0.66

Tokens

-0.64

place

-0.62

ier

-0.61

 grate

-0.60

arro

-0.59

POSITIVE LOGITS

iciary

0.75

 household

0.74

intelligence

0.74

 staff

0.74

CHQ

0.73

 Household

0.72

 state

0.71

 overseeing

0.71

 department

0.70

staff

0.70

Activations Density 0.095%

titles or positions related to leadership or authority

phrases containing the term "head of" in various contexts related to authority or leadership positions

No Comments

No Known Activations