INDEX

Explanations

proper nouns, specifically related to the name "El Capitan"

oai_token-act-pair · gpt-3.5-turbo

references to the word "Titan" and its related forms

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

pressure

-0.78

EStreamFrame

-0.68

allery

-0.67

displayText

-0.66

Trend

-0.66

oney

-0.65

 projected

-0.64

score

-0.64

 Dise

-0.63

anwhile

-0.62

POSITIVE LOGITS

itan

1.44

osaurs

1.00

igans

0.95

itans

0.94

ica

0.91

iac

0.88

ibles

0.87

ic

0.87

igan

0.84

auts

0.84

Activations Density 0.002%

proper nouns, specifically related to the name "El Capitan"

references to the word "Titan" and its related forms

No Comments

No Known Activations