INDEX

Explanations

things that are restricted or not allowed beyond a certain point

oai_token-act-pair · gpt-3.5-turbo

instances of the word "limited."

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

ayers

-0.72

ayer

-0.72

ortmund

-0.72

frog

-0.66

agles

-0.66

evil

-0.65

faced

-0.65

pointer

-0.64

 exclaim

-0.64

eton

-0.63

POSITIVE LOGITS

 limited

3.59

limited

2.51

Limited

2.33

 LIMITED

2.30

 restricted

2.11

 Limited

2.06

 constrained

1.74

 confined

1.63

 limitations

1.60

 limiting

1.58

Activations Density 0.028%

things that are restricted or not allowed beyond a certain point

instances of the word "limited."

No Comments

No Known Activations