INDEX

Explanations

references to nobility titles, especially the title "Lord."

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Lord."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

¥ŀ

-0.90

Accessory

-0.74

 cellul

-0.72

ickr

-0.70

ascus

-0.69

insula

-0.68

OPLE

-0.68

utic

-0.66

single

-0.66

 conduc

-0.66

POSITIVE LOGITS

lord

1.02

Lord

0.96

ess

0.92

friend

0.86

nard

0.86

 Jesus

0.85

esses

0.85

frey

0.84

stown

0.84

 Lord

0.84

Activations Density 0.005%

references to nobility titles, especially the title "Lord."

mentions of the term "Lord."

No Comments

No Known Activations