INDEX

Explanations

proper nouns and titles related to nobility and fantasy settings

oai_token-act-pair · gpt-3.5-turbo

references to the "Lord of the Rings" series

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

*/(

-0.76

inarily

-0.76

ILY

-0.72

 partName

-0.71

BILL

-0.70

UGC

-0.69

culosis

-0.68

wcs

-0.67

uffy

-0.65

paren

-0.65

POSITIVE LOGITS

 Souls

0.74

 Persia

0.74

 course

0.74

Wak

0.72

icial

0.72

 Atlantis

0.71

 sorts

0.71

 whom

0.69

 renown

0.69

 Rome

0.68

Activations Density 0.179%

proper nouns and titles related to nobility and fantasy settings

references to the "Lord of the Rings" series

No Comments

No Known Activations