INDEX

Explanations

phrases related to legal or judicial actions

oai_token-act-pair · gpt-3.5-turbo

references to physical punishment or legal consequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 proble

-0.72

itiz

-0.71

pires

-0.68

Ö¼

-0.62

{*

-0.60

guyen

-0.57

inventoryQuantity

-0.57

ricular

-0.56

ãĥĺ

-0.55

kie

-0.54

POSITIVE LOGITS

 respectively

2.42

 apiece

1.98

 together

1.45

 respective

1.42

 themselves

1.26

selves

1.22

 jointly

1.21

 collectively

1.20

together

1.11

 each

1.09

Activations Density 0.678%

phrases related to legal or judicial actions

references to physical punishment or legal consequences

No Comments

No Known Activations