INDEX

Explanations

references to the Pope or Vatican

oai_token-act-pair · gpt-3.5-turbo

mentions of the Pope and related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

UCT

-0.79

hirt

-0.72

IDES

-0.70

é¾

-0.66

VE

-0.66

IGHTS

-0.65

CONCLUS

-0.64

 learners

-0.64

Interstitial

-0.64

 Supplemental

-0.63

POSITIVE LOGITS

yrus

1.48

rika

1.24

illon

1.06

arella

1.06

ilion

1.06

illion

1.05

aya

1.01

oleon

1.01

ular

0.99

ersen

0.99

Activations Density 0.030%

references to the Pope or Vatican

mentions of the Pope and related terminology

No Comments

No Known Activations