INDEX

Explanations

words related to expressing strong emotions, particularly shouting or screaming

oai_token-act-pair · gpt-3.5-turbo

instances of laughter or screaming in contexts of humor and emotional expression

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

Not in Any Lists

No Comments

Negative Logits

qua

-0.77

ngth

-0.77

ormal

-0.73

patrick

-0.71

iterranean

-0.70

chnology

-0.69

Else

-0.68

idious

-0.67

levard

-0.67

oteric

-0.66

POSITIVE LOGITS

 loud

1.19

 louder

1.18

 loudly

1.11

 uncontroll

1.02

 inco

0.96

 plaint

0.85

 laughter

0.85

 helpless

0.84

 aloud

0.84

 hysterical

0.84

Activations Density 0.378%

words related to expressing strong emotions, particularly shouting or screaming

instances of laughter or screaming in contexts of humor and emotional expression

No Comments

No Known Activations