INDEX

Explanations

patterns related to formatting and structure in text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 Geſch

-0.79

 Infórmanos

-0.78

Diweddarwch

-0.74

 المعيارى

-0.72

 للمعارف

-0.71

ArrowToggle

-0.71

 Normdatei

-0.71

Ӕ

-0.70

delwed

-0.69

 يتيمه

-0.69

POSITIVE LOGITS

Un

0.46

This

0.45

Un

0.45

The

0.44

0.41

Today

0.41

Self

0.40

<eos>

0.39

0.37

Activations Density 1.659%

patterns related to formatting and structure in text

No Comments

No Known Activations