INDEX

Explanations

proper nouns of people or entities appearing in media contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "appeared" in various contexts

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

dp

-0.76

brace

-0.70

rpm

-0.67

 vines

-0.67

ribution

-0.66

EStreamFrame

-0.65

current

-0.65

venant

-0.64

ricular

-0.64

currency

-0.63

POSITIVE LOGITS

Pause

0.80

 HuffPost

0.66

 Minion

0.66

 Fargo

0.64

 Cipher

0.64

Assistant

0.63

 appearances

0.62

imony

0.61

 onstage

0.60

lihood

0.60

Activations Density 0.826%

proper nouns of people or entities appearing in media contexts

instances of the word "appeared" in various contexts

No Comments

No Known Activations