INDEX

Explanations

mentions of proper nouns, particularly names starting with "Sp"

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Sp."

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

ãĥĥãĥĪ

-0.77

DragonMagazine

-0.71

ographed

-0.71

cipline

-0.70

*/(

-0.68

riad

-0.67

ģĸ

-0.66

İĭ

-0.65

cised

-0.65

tical

-0.64

POSITIVE LOGITS

aghetti

0.99

orp

0.97

itt

0.96

atial

0.92

irts

0.90

inn

0.84

Sp

0.84

acious

0.83

encer

0.83

elt

0.83

Activations Density 0.007%

mentions of proper nouns, particularly names starting with "Sp"

mentions of the term "Sp."

No Comments

No Known Activations