INDEX

Explanations

references to customization or configuration processes

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

custom communication channel

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

 Administrativna

-0.79

MarshalTo

-0.63

 kaarangay

-0.61

⅘

-0.57

TypedDataSet

-0.57

 صوتيه

-0.54

 betweenstory

-0.54

PMailer

-0.53

RegressionTest

-0.51

 propOrder

-0.51

POSITIVE LOGITS

">//

0.49

httphttps

0.47

]]

0.36

 uska

0.32

 ſur

0.32

"];

0.32

0.31

 irrational

0.30

 sandero

0.29

 juſ

0.29

Activations Density 0.000%

references to customization or configuration processes

custom communication channel

No Comments

No Known Activations