INDEX

Explanations

special characters or symbols indicating a new section or break in the text

oai_token-act-pair · gpt-3.5-turbo

instances of significant numerical or statistical data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 utilizing

-0.88

 favorably

-0.86

 honored

-0.86

 utilized

-0.85

 totaling

-0.82

 Hawaiian

-0.79

 afterward

-0.78

 counselors

-0.78

 honors

-0.77

 honoring

-0.77

POSITIVE LOGITS

alys

1.23

 Whilst

1.22

BBC

1.20

Shape

1.19

Labour

1.17

 analyse

1.16

 realise

1.15

Britain

1.14

 (Â£

1.13

 Â£

1.12

Activations Density 0.463%

special characters or symbols indicating a new section or break in the text

instances of significant numerical or statistical data

No Comments

No Known Activations