INDEX

Explanations

mentions of "Federal" entities

oai_token-act-pair · gpt-3.5-turbo

references to federal entities or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 arte

-0.75

 whis

-0.69

aphael

-0.67

awa

-0.66

 sprite

-0.65

mouse

-0.64

 sponge

-0.64

 palette

-0.64

 sweets

-0.63

POV

-0.62

POSITIVE LOGITS

 Federal

3.62

Federal

3.12

 federal

2.29

ederal

1.76

Fed

1.61

 Government

1.59

 feds

1.58

 Presidential

1.54

 federally

1.52

 Congressional

1.52

Activations Density 0.015%

mentions of "Federal" entities

references to federal entities or organizations

No Comments

No Known Activations