INDEX

Explanations

elements related to data structures or programming constructs

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

lists followed by punctuation or index

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

-0.57

Localisation

-0.40

 thiệu

-0.40

posedge

-0.36

 nationalen

-0.35

Typeface

-0.35

 defStyle

-0.34

 Statue

-0.33

localisation

-0.33

Statue

-0.33

POSITIVE LOGITS

SizeMode

0.55

ⓧ

0.53

 ujednoznacz

0.51

XmlEnum

0.49

 بيها

0.48

ModelSerializer

0.48

msgTypes

0.47

のうち

0.47

 utafitiHapana

0.46

[]*

0.46

Activations Density 0.017%

elements related to data structures or programming constructs

lists followed by punctuation or index

No Comments

No Known Activations