INDEX

Explanations

references to individuals in leadership or key roles within organizations

oai_token-act-pair · gpt-3.5-turbo

the term "showrunner" and its variations in context to television or film production

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

olia

-0.79

ores

-0.70

enta

-0.69

berra

-0.68

ensed

-0.67

ork

-0.67

 minors

-0.65

oret

-0.65

 plight

-0.64

olk

-0.64

POSITIVE LOGITS

runner

1.39

runners

1.10

ners

0.86

swick

0.83

 Runner

0.83

stellar

0.82

Runner

0.80

 runner

0.79

backer

0.79

luster

0.77

Activations Density 0.004%

references to individuals in leadership or key roles within organizations

the term "showrunner" and its variations in context to television or film production

No Comments

No Known Activations