INDEX

Explanations

locations or places

oai_token-act-pair · gpt-3.5-turbo

references to locations or settings within a narrative

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

govtrack

-0.85

-0.74

Secondly

-0.74

 Secondly

-0.73

bear

-0.73

thood

-0.73

 secondly

-0.73

uz

-0.71

erest

-0.70

manship

-0.69

POSITIVE LOGITS

 rooft

1.44

 streets

1.43

 countryside

1.42

 halls

1.37

 battlefield

1.29

 floors

1.24

 landscape

1.22

 skies

1.21

 corridors

1.19

 globe

1.19

Activations Density 0.180%

locations or places

references to locations or settings within a narrative

No Comments

No Known Activations