INDEX

Explanations

proper nouns and names associated with specific entities or places

oai_token-act-pair · gpt-3.5-turbo

names and mentions of individuals, particularly those associated with specific stories or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 smoker

-0.77

========

-0.71

iaries

-0.68

hetically

-0.68

wered

-0.67

¥ŀ

-0.67

indust

-0.67

oult

-0.66

avorite

-0.66

ictions

-0.66

POSITIVE LOGITS

 Ness

0.92

 Rhodes

0.81

 Loch

0.80

Afee

0.78

gow

0.77

inosaur

0.76

poke

0.75

 Colossus

0.75

bones

0.74

Cro

0.74

Activations Density 0.023%

proper nouns and names associated with specific entities or places

names and mentions of individuals, particularly those associated with specific stories or events

No Comments

No Known Activations