INDEX

Explanations

words with the suffix "-end"

oai_token-act-pair · gpt-3.5-turbo

references to high-tech advancements or devices

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

ulhu

-0.79

 Creator

-0.72

ļéĨĴ

-0.67

 Dull

-0.67

Cah

-0.66

 Borders

-0.66

ħĭ

-0.65

 Moreno

-0.65

 Richards

-0.64

 Norris

-0.63

POSITIVE LOGITS

level

1.23

powered

1.11

sounding

1.09

sized

1.08

grade

1.07

ranking

1.06

eyed

1.01

interest

1.00

performing

1.00

enough

0.99

Activations Density 0.047%

words with the suffix "-end"

references to high-tech advancements or devices

No Comments

No Known Activations