INDEX

Explanations

mathematical expressions and calculations

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

function definitions using parentheses

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

 siden

-0.42

 ministry

-0.32

 Schoen

-0.32

 Hess

-0.32

 THINK

-0.31

 Brust

-0.31

と思ったら

-0.31

 side

-0.30

 think

-0.30

 morale

-0.30

POSITIVE LOGITS

OGND

0.78

 okuyayım

0.69

 ModelExpression

0.60

CppMethod

0.60

PostExecute

0.60

참고

0.56

Derbyniad

0.55

PostInfinity

0.54

ScopeManager

0.54

RefNanny

0.54

Activations Density 0.595%

mathematical expressions and calculations

function definitions using parentheses

No Comments

No Known Activations