INDEX

Explanations

key arguments in a problem or code context

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

conjunctions and common words

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

:✨

-0.57

casian

-0.48

DIPSETTING

-0.43

 Submarine

-0.43

ASE

-0.43

 Transistor

-0.41

WebVitals

-0.40

createNewFile

-0.40

évaluateur

-0.40

charest

-0.40

POSITIVE LOGITS

 propOrder

0.53

IsContent

0.50

BagLayout

0.49

enumii

0.45

tagHelperRunner

0.44

 تضيفلها

0.43

isContained

0.43

WithIOException

0.42

 esternos

0.42

 <=",

0.42

Activations Density 0.129%

key arguments in a problem or code context

conjunctions and common words

No Comments

No Known Activations