INDEX

Explanations

words related to influence or power

oai_token-act-pair · gpt-3.5-turbo

references to influence in various contexts

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

ft

-0.79

eared

-0.75

DEF

-0.74

asar

-0.72

adapt

-0.70

leigh

-0.69

 Dill

-0.68

ears

-0.66

onew

-0.64

York

-0.64

POSITIVE LOGITS

 pedd

1.20

 exerted

1.16

 influencing

1.03

 shaping

0.95

 exercised

0.92

 influence

0.92

 sway

0.86

 cooker

0.83

 wielded

0.83

 influences

0.80

Activations Density 0.050%

words related to influence or power

references to influence in various contexts

No Comments

No Known Activations