INDEX

Explanations

adjectives related to strong emotions, particularly anger

oai_token-act-pair · gpt-3.5-turbo

instances of the word "mad" 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 @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

•layer 8 - anger

No Comments

Negative Logits

sites

-0.74

utical

-0.71

 Vert

-0.66

ãģ¯

-0.65

 Islands

-0.64

é¾įå¥ĳå£«

-0.63

phrine

-0.61

etitive

-0.61

 Graveyard

-0.60

soDeliveryDate

-0.59

POSITIVE LOGITS

rid

1.08

agascar

0.96

der

0.91

cap

0.87

rup

0.87

iated

0.81

cow

0.80

ly

0.80

oline

0.80

rams

0.79

Activations Density 0.028%

adjectives related to strong emotions, particularly anger

instances of the word "mad" in various contexts

No Comments

No Known Activations