INDEX

Explanations

mentions of the word "President" or related terms

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Pres" or references to presidential figures

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Lions

-0.74

STAR

-0.70

 brackets

-0.69

 Abyssal

-0.67

 subt

-0.67

EStreamFrame

-0.65

ļéĨĴ

-0.63

 Chop

-0.62

 loopholes

-0.61

 Salvation

-0.61

POSITIVE LOGITS

cription

1.19

chool

1.16

ervation

1.11

umed

1.10

erves

1.07

idency

1.06

byter

1.05

ervative

1.04

umably

1.04

erving

1.04

Activations Density 0.008%

mentions of the word "President" or related terms

mentions of the word "Pres" or references to presidential figures

No Comments

No Known Activations