INDEX

Explanations

words containing non-English characters, specifically umlauts and other accents

oai_token-act-pair · gpt-3.5-turbo

specific names and terms associated with individuals and locations

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

 Debor

-0.83

 Fired

-0.73

 Reef

-0.73

 Rhino

-0.69

 scorp

-0.68

ORK

-0.67

 Turtles

-0.66

BILITIES

-0.65

yrinth

-0.64

ombat

-0.64

POSITIVE LOGITS

Ã¶n

1.44

vironment

1.00

ning

0.85

Ã¼

0.83

bach

0.81

Ã¤

0.79

icht

0.79

kamp

0.77

agar

0.76

ollen

0.76

Activations Density 0.008%

words containing non-English characters, specifically umlauts and other accents

specific names and terms associated with individuals and locations

No Comments

No Known Activations