INDEX

Explanations

references to antagonists in fictional stories

oai_token-act-pair · gpt-3.5-turbo

references to villains in narratives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

RT

-0.79

kt

-0.77

Input

-0.76

FM

-0.73

FM

-0.71

CLS

-0.71

IGHTS

-0.71

parts

-0.70

Mat

-0.69

ounds

-0.68

POSITIVE LOGITS

 villain

3.33

 villains

2.63

 antagonist

1.59

 traitor

1.52

vill

1.30

 coward

1.26

 heroine

1.19

 trope

1.16

 treacher

1.15

 philanthrop

1.11

Activations Density 0.029%

references to antagonists in fictional stories

references to villains in narratives

No Comments

No Known Activations