INDEX

Explanations

key phrases and transitions in written content that indicate importance or emphasis in discussions or FAQs

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

information, reviews, or descriptions

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

TagMode

-0.61

MigrationBuilder

-0.57

PreExecute

-0.57

الإنجليزية

-0.56

 propOrder

-0.54

ágenes

-0.54

 ProtoMessage

-0.53

 محفوظة

-0.52

inSlope

-0.52

 InputDecoration

-0.52

POSITIVE LOGITS

Love

0.39

 veggie

0.38

 smashed

0.38

 fast

0.38

lifer

0.38

[...]

0.38

 gorgeous

0.37

 love

0.37

uns

0.36

 insanely

0.36

Activations Density 0.009%

key phrases and transitions in written content that indicate importance or emphasis in discussions or FAQs

information, reviews, or descriptions

No Comments

No Known Activations