INDEX

Explanations

phrases indicating challenges or problematic situations

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

states and outcomes

np_acts-logits-general · gemini-2.5-flash-lite

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

 للمعارف

-0.72

 Chwiliwch

-0.72

 autorytatywna

-0.67

Hentet

-0.66

Билгалдахарш

-0.65

<unused14>

-0.64

<unused68>

-0.64

<pad>

-0.64

AndEndTag

-0.64

[@BOS@]

-0.63

POSITIVE LOGITS

rsiniz

0.33

 peuples

0.32

wä

0.32

jatel

0.32

 pretty

0.32

 становника

0.30

GO

0.29

 pouvoirs

0.28

Activations Density 0.102%

phrases indicating challenges or problematic situations

states and outcomes

No Comments

No Known Activations