INDEX

Explanations

personal pronouns followed by actions or attributes

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "she."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

arius

-0.63

achev

-0.58

ãĥ¯

-0.58

eco

-0.57

respective

-0.54

resents

-0.54

 Software

-0.50

çīĪ

-0.50

 âĺħ

-0.49

iPhone

-0.49

POSITIVE LOGITS

she

3.18

she

2.65

She

2.42

SHE

2.25

She

2.09

her

1.93

 hers

1.90

 herself

1.69

Her

1.56

Her

1.37

Activations Density 0.080%

personal pronouns followed by actions or attributes

occurrences of the word "she."

No Comments

No Known Activations