INDEX

Explanations

references to theater and performance-related activities

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

theater, opera, drama

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

 sense

-0.34

-0.31

 there

-0.28

 défa

-0.26

 physical

-0.26

 There

-0.25

fa

-0.25

 primary

-0.25

mõ

-0.25

↵↵

-0.24

POSITIVE LOGITS

+#+

0.84

ſelf

0.80

 betweenstory

0.76

NameInMap

0.75

 ſind

0.75

 erſt

0.73

ImageContext

0.73

<unused74>

0.73

<pad>

0.73

[@BOS@]

0.73

Activations Density 0.002%

references to theater and performance-related activities

theater, opera, drama

No Comments

No Known Activations