INDEX

Explanations

keywords and terms related to specific technical processes and concepts

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

concept followed by specifier

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

LookAnd

-0.62

#+#

-0.56

+#+#

-0.54

transQ

-0.53

 JpaRepository

-0.52

 nonUne

-0.52

AndEndTag

-0.51

 noDo

-0.50

UnusedPrivate

-0.50

HomeAsUpEnabled

-0.49

POSITIVE LOGITS

 nahilalakip

0.55

رشف

0.40

 esternos

0.36

multicolumn

0.35

asci

0.33

'{@

0.33

">//

0.33

AxisSize

0.33

 gynhyrchwyd

0.33

 cherchés

0.32

Activations Density 0.045%

keywords and terms related to specific technical processes and concepts

concept followed by specifier

No Comments

No Known Activations