INDEX

Explanations

references to variables or fields in a programming context

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

m followed by variable names

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

存于互联网档案馆

-0.72

XMLSchema

-0.63

featureID

-0.59

InSection

-0.59

HtmlAttribute

-0.56

verifyException

-0.55

typeorm

-0.55

DoubleQuotes

-0.55

numerusform

-0.54

Πηγές

-0.54

POSITIVE LOGITS

 Minang

0.40

 defaultstate

0.36

TintMode

0.36

rugada

0.35

 CONSUMER

0.34

 kapture

0.34

 dezelve

0.33

 meus

0.33

Nonnull

0.33

 Bucure

0.33

Activations Density 0.005%

references to variables or fields in a programming context

m followed by variable names

No Comments

No Known Activations