INDEX

Explanations

phrases or words related to interactions with or actions towards a specific female character

oai_token-act-pair · gpt-3.5-turbo

references to a specific female subject

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ype

-1.00

ypes

-0.93

Ĥª

-0.77

ornia

-0.77

arios

-0.76

hovah

-0.75

kefeller

-0.74

anamo

-0.74

ffic

-0.73

ormons

-0.72

POSITIVE LOGITS

 husband

1.41

 daughter

1.35

 granddaughter

1.25

 niece

1.23

 boyfriend

1.19

 sister

1.16

 breasts

1.12

 daughters

1.12

 purse

1.08

 grandson

1.08

Activations Density 0.098%

phrases or words related to interactions with or actions towards a specific female character

references to a specific female subject

No Comments

No Known Activations