INDEX

Explanations

spoilers in a text

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate the presence of spoilers in a text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

kt

-0.73

ndra

-0.73

urat

-0.69

rique

-0.69

leanor

-0.68

trak

-0.67

ŃĶ

-0.67

riad

-0.67

ton

-0.66

¬¼

-0.66

POSITIVE LOGITS

 spoilers

1.17

 spoiler

1.16

oiler

1.01

 Spoiler

0.95

OIL

0.94

spo

0.93

 spoil

0.92

ervative

0.83

 Collider

0.75

":""},{"

0.75

Activations Density 0.031%

spoilers in a text

phrases that indicate the presence of spoilers in a text

No Comments

No Known Activations