INDEX

Explanations

punctuation marks and formatting elements in written text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

 betweenstory

-0.85

ThroughAttribute

-0.77

<unused23>

-0.72

[@BOS@]

-0.72

<unused41>

-0.72

<unused14>

-0.72

<unused28>

-0.72

<unused17>

-0.72

<unused3>

-0.72

<unused8>

-0.72

POSITIVE LOGITS

ritch

0.30

 Mindest

0.27

HasAnnotation

0.27

WSER

0.26

strona

0.26

 jeszcze

0.25

 sayfası

0.25

 chưa

0.24

입

0.24

 sentarse

0.24

Activations Density 0.000%

punctuation marks and formatting elements in written text

No Comments

No Known Activations