INDEX

Explanations

the word "Hero" with various intensity levels

oai_token-act-pair · gpt-3.5-turbo

references to the term "Hero" and its variations

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

ateur

-0.77

aeda

-0.76

choes

-0.76

creen

-0.74

cling

-0.67

model

-0.67

nce

-0.66

FORMATION

-0.66

acular

-0.66

ĸļ

-0.64

POSITIVE LOGITS

Hero

1.19

 Hero

1.02

 Chaser

0.78

ines

0.77

inous

0.76

inic

0.76

 heroine

0.73

esses

0.71

uin

0.71

 Soldier

0.70

Activations Density 0.013%

the word "Hero" with various intensity levels

references to the term "Hero" and its variations

No Comments

No Known Activations