INDEX

Explanations

references to operating system and related code directories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iriman

-0.35

Werdegang

-0.35

じん

-0.31

 fogo

-0.30

 Foire

-0.30

linspace

-0.29

oyl

-0.29

Cog

-0.29

 Tuck

-0.29

 circum

-0.28

POSITIVE LOGITS

Os

0.73

protoimpl

0.68

os

0.65

 Osbourne

0.65

Autoritní

0.62

Os

0.62

OSC

0.59

IntoConstraints

0.57

 oscar

0.57

OS

0.56

Activations Density 0.127%

references to operating system and related code directories

No Comments

No Known Activations