INDEX

Explanations

references to bodily parts and actions specifically related to throats and mouths

oai_token-act-pair · gpt-3.5-turbo

references to violence and body parts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

gran

-0.60

Ping

-0.60

 massage

-0.59

 therapy

-0.58

port

-0.58

 guru

-0.57

 character

-0.56

orient

-0.56

 therapist

-0.55

gum

-0.55

POSITIVE LOGITS

 throats

2.73

 mouths

1.49

 necks

1.10

 births

1.08

aments

0.85

 noses

0.80

 asses

0.77

 Pigs

0.75

 lives

0.73

 taxp

0.72

Activations Density 0.010%

references to bodily parts and actions specifically related to throats and mouths

references to violence and body parts

No Comments

No Known Activations