INDEX

Explanations

references to newness or novelty

oai_token-act-pair · gpt-3.5-turbo

references to the concept of being "new."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

acebook

-0.81

aminer

-0.81

iage

-0.74

istine

-0.72

AFTA

-0.70

itta

-0.69

ammers

-0.69

UMP

-0.67

OPA

-0.66

icone

-0.66

POSITIVE LOGITS

bie

1.42

bies

1.41

 Zealand

1.10

foundland

0.98

egg

0.85

born

0.85

 arrivals

0.85

 Orleans

0.80

 Testament

0.80

found

0.80

Activations Density 0.114%

references to newness or novelty

references to the concept of being "new."

No Comments

No Known Activations