INDEX

Explanations

elements conveying emphasis or importance in context

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

questioning others

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

MessageTagHelper

-0.68

IsMutable

-0.67

GEBURTSDATUM

-0.67

 ExecuteAsync

-0.67

 AssemblyCulture

-0.66

 beginnetje

-0.63

 protoimpl

-0.62

EDEFAULT

-0.61

Publica

-0.59

 poveznice

-0.59

POSITIVE LOGITS

FXML

0.31

 interviewed

0.30

born

0.29

 réaction

0.28

 spoken

0.28

behind

0.27

leader

0.27

Boot

0.27

LEAD

0.27

ANTLR

0.27

Activations Density 0.005%

elements conveying emphasis or importance in context

questioning others

No Comments

No Known Activations