INDEX

Explanations

specific terms and labels in a categorized or structured list format

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

anomalous lines, odd row, females, malabsorption

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

RefNanny

-0.63

 صوتيه

-0.59

+#+#

-0.56

ScopeManager

-0.54

 الحره

-0.52

 utafitiHapana

-0.52

rungsseite

-0.52

tagHelperRunner

-0.51

nail

-0.50

GOTREF

-0.49

POSITIVE LOGITS

tinyos

0.38

propTypes

0.36

 jsPsych

0.36

devServer

0.36

 bArr

0.36

wapV

0.35

PreferredItem

0.35

WithIdentifier

0.34

 also

0.34

ruptedException

0.34

Activations Density 0.126%

specific terms and labels in a categorized or structured list format

anomalous lines, odd row, females, malabsorption

No Comments

No Known Activations