INDEX

Explanations

temporal markers and specific dates related to events or actions

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

years and dates

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

Tikang

-0.48

Bakgrunnsstoff

-0.43

 CommonModule

-0.43

COUVER

-0.42

 stiff

-0.41

ColumnHeaders

-0.41

 betweenstory

-0.39

 cenar

-0.38

 esist

-0.37

ediakan

-0.37

POSITIVE LOGITS

 ویکی‌پدیای

0.43

SerializedSize

0.41

 appunto

0.40

所以

0.39

GOTREF

0.38

EndContext

0.37

ClientSize

0.37

parsedMessage

0.36

 oprot

0.36

 تكبرها

0.36

Activations Density 0.028%

temporal markers and specific dates related to events or actions

years and dates

No Comments

No Known Activations