INDEX

Explanations

references to religious or sacred places, specifically shrines

oai_token-act-pair · gpt-3.5-turbo

references to shrines and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Winchester

-0.71

eering

-0.71

 Gutenberg

-0.70

lace

-0.68

POL

-0.68

 Hebdo

-0.68

 Prospect

-0.67

 Trojan

-0.64

 laundering

-0.64

 PowerPoint

-0.64

POSITIVE LOGITS

ines

0.97

shr

0.96

inks

0.95

anch

0.91

ower

0.91

iffs

0.91

abies

0.90

otted

0.89

Ń·

0.89

ubs

0.84

Activations Density 0.008%

references to religious or sacred places, specifically shrines

references to shrines and related terms

No Comments

No Known Activations