INDEX

Explanations

names related to legal cases or situations

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals and terms associated with crime and injury

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Â¢

-0.82

sonian

-0.79

REC

-0.69

icz

-0.68

racuse

-0.68

atcher

-0.64

WT

-0.63

ĨĴ

-0.63

\":

-0.63

POW

-0.63

POSITIVE LOGITS

aceae

0.85

bia

0.82

lip

0.79

butt

0.76

galitarian

0.76

icide

0.75

 Ventura

0.72

neau

0.72

alore

0.70

bag

0.68

Activations Density 0.029%

names related to legal cases or situations

references to specific individuals and terms associated with crime and injury

No Comments

No Known Activations