INDEX

Explanations

specific mentions of courthouses in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to court settings or judicial contexts

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

ORN

-0.61

 weights

-0.61

 Franks

-0.56

 Burgess

-0.55

 extreme

-0.55

 heavy

-0.55

 curfew

-0.54

 Shining

-0.54

 remotely

-0.54

 Stafford

-0.54

POSITIVE LOGITS

tyard

1.74

thouse

1.45

tesy

1.44

ser

1.40

sers

1.33

sing

1.32

ses

1.32

ten

1.28

sed

1.22

tes

1.20

Activations Density 0.057%

specific mentions of courthouses in various contexts

references to court settings or judicial contexts

No Comments

No Known Activations