INDEX

Explanations

movie-related terms

oai_token-act-pair · gpt-3.5-turbo

references to movies or media content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

vana

-0.73

picking

-0.72

reth

-0.71

picked

-0.71

aez

-0.67

©¶æ

-0.65

izontal

-0.65

attery

-0.65

agher

-0.64

ber

-0.62

POSITIVE LOGITS

zx

0.78

ension

0.78

ensions

0.76

 includ

0.75

EMENT

0.72

tsky

0.71

ments

0.70

ISION

0.69

EMBER

0.68

============

0.67

Activations Density 0.022%

movie-related terms

references to movies or media content

No Comments

No Known Activations