INDEX

Explanations

legal and law enforcement-related terms

oai_token-act-pair · gpt-3.5-turbo

mentions of criminal activities or incidents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Todd

-0.76

 Garrett

-0.72

Ohio

-0.69

 Shelby

-0.68

 Gotham

-0.68

 Dungeons

-0.68

DeL

-0.67

 Slate

-0.67

 Whitney

-0.67

 Lovecraft

-0.67

POSITIVE LOGITS

 Malays

0.93

 Tamil

0.90

 Kerala

0.87

 learnt

0.87

 Hindi

0.85

 authorised

0.85

arij

0.84

 Modi

0.83

 organis

0.83

jriwal

0.83

Activations Density 3.902%

legal and law enforcement-related terms

mentions of criminal activities or incidents

No Comments

No Known Activations