INDEX

Explanations

legal and law enforcement related terms and actions

oai_token-act-pair · gpt-3.5-turbo

words associated with legal actions and judicial processes

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

Interstitial

-0.81

 teasp

-0.81

£ı

-0.75

 exting

-0.68

zbollah

-0.67

ļéĨĴ

-0.66

ongyang

-0.65

 helicop

-0.65

Nintendo

-0.62

Invalid

-0.62

POSITIVE LOGITS

him

1.92

him

1.18

 them

0.98

HIM

0.98

Mr

0.94

Him

0.90

 Meadows

0.83

her

0.80

Ms

0.76

 Cullen

0.75

Activations Density 0.517%

legal and law enforcement related terms and actions

words associated with legal actions and judicial processes

No Comments

No Known Activations