INDEX

Explanations

programming-related output and logging statements

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

commands and code execution

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

 '\\;'

-0.97

 الرياضيه

-0.88

 Normdatei

-0.88

 BoxDecoration

-0.87

 zuſammen

-0.87

Personensuche

-0.86

 invokingState

-0.83

AndEndTag

-0.81

complexType

-0.80

<unused41>

-0.79

POSITIVE LOGITS

 immediately

0.27

自

0.27

ч

0.26

0.25

 small

0.25

 afec

0.25

 then

0.25

CONNECTION

0.24

Immediately

0.24

 weißem

0.23

Activations Density 0.012%

programming-related output and logging statements

commands and code execution

No Comments

No Known Activations

programming-related output and logging statements

commands and code execution

No Comments

No Known Activations