INDEX

Explanations

positions or roles of authority within specific organizations

oai_token-act-pair · gpt-3.5-turbo

terms associated with leadership roles and responsibilities

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

behind

-0.81

nery

-0.77

ascript

-0.76

peak

-0.76

hooting

-0.75

dating

-0.74

ternity

-0.74

 reacting

-0.74

seeing

-0.72

picking

-0.70

POSITIVE LOGITS

 factors

0.80

 heights

0.78

 factor

0.77

 Factors

0.75

 edge

0.75

 philosophies

0.73

 force

0.71

åĬ

0.70

 professions

0.70

iaries

0.70

Activations Density 0.173%

positions or roles of authority within specific organizations

terms associated with leadership roles and responsibilities

No Comments

No Known Activations