INDEX

Explanations

proper nouns, specifically names of people and places

oai_token-act-pair · gpt-3.5-turbo

names of places and people, particularly in a geographical context

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

cer

-0.78

Americ

-0.77

itzer

-0.74

cific

-0.74

onies

-0.70

dan

-0.69

CVE

-0.68

Iraq

-0.67

alog

-0.65

zees

-0.65

POSITIVE LOGITS

 Everett

1.08

verett

0.89

cffff

0.81

 Tacoma

0.81

 Dickinson

0.81

 Lowell

0.81

é¾įå

0.78

å§«

0.74

ITNESS

0.74

 Mariners

0.73

Activations Density 0.016%

proper nouns, specifically names of people and places

names of places and people, particularly in a geographical context

No Comments

No Known Activations