INDEX

Explanations

equal signs likely related to code or configuration settings

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

identifier equals value

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

IUrlHelper

-0.59

$_"

-0.59

hdashline

-0.58

 ſta

-0.57

expandindo

-0.57

 ſch

-0.56

存于互联网档案馆

-0.54

ſelf

-0.54

({"

-0.53

 Houſe

-0.53

POSITIVE LOGITS

1.23

0.76

>=</

0.72

$=

0.71

=\

0.69

:=

0.68

=\

0.68

==

0.67

$=\

0.66

}=

0.64

Activations Density 0.031%

equal signs likely related to code or configuration settings

identifier equals value

No Comments

No Known Activations