INDEX

Explanations

locations or places

oai_token-act-pair · gpt-3.5-turbo

references to a specific entity or concept identified as "Pl."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

âĸ¬âĸ¬

-0.78

 Logged

-0.76

worthiness

-0.68

ãģ®éŃĶ

-0.65

ãĥĨãĤ£

-0.64

âĸ¬

-0.63

 nexus

-0.63

 liking

-0.60

 FANTASY

-0.60

manship

-0.59

POSITIVE LOGITS

umbing

1.42

atoon

1.40

utonium

1.37

acement

1.29

ummer

1.27

ausible

1.23

acid

1.22

atter

1.22

umb

1.20

acer

1.20

Activations Density 0.011%

locations or places

references to a specific entity or concept identified as "Pl."

No Comments

No Known Activations