INDEX

Explanations

references to prototypes and prototyping processes

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

prototype and early versions

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

farande

-0.38

bės

-0.33

 peuple

-0.32

 peaceful

-0.30

Tod

-0.30

레스

-0.29

<-

-0.29

<eos>

-0.29

İstinadlar

-0.28

 precin

-0.28

POSITIVE LOGITS

 createSprite

0.89

parsedMessage

0.79

twimg

0.74

adaptiveStyles

0.67

tagext

0.64

MigrationBuilder

0.62

richTextPanel

0.62

WithIOException

0.60

 <>",

0.60

principalColumn

0.59

Activations Density 0.024%

references to prototypes and prototyping processes

prototype and early versions

No Comments

No Known Activations