INDEX

Explanations

words related to utilizing power or influence for various purposes

oai_token-act-pair · gpt-3.5-turbo

mentions of power and 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 @ 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

ategories

-0.76

:=

-0.70

ods

-0.68

%:

-0.67

raid

-0.66

aughtered

-0.65

licts

-0.65

itant

-0.65

ovy

-0.64

%]

-0.63

POSITIVE LOGITS

 playbook

0.77

 bluff

0.76

 expertise

0.74

 sparing

0.73

 pseudonym

0.72

 loophole

0.69

 analogy

0.69

 clout

0.68

 liber

0.67

 metaphor

0.67

Activations Density 0.422%

words related to utilizing power or influence for various purposes

mentions of power and influence in various contexts

No Comments

No Known Activations