INDEX

Explanations

curly braces and structure markers in code

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

code formatting and keywords

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

coma

-0.47

rancy

-0.44

irability

-0.44

granate

-0.44

nades

-0.43

ofed

-0.42

mable

-0.42

limia

-0.42

drew

-0.42

tituted

-0.41

POSITIVE LOGITS

 ویکی‌پدی

0.58

 initComponents

0.50

 betweenstory

0.50

 AssemblyCulture

0.49

 endwhile

0.46

drawiam

0.45

 propOrder

0.44

 ivelany

0.42

 '\\;'

0.42

oredCriteria

0.41

Activations Density 0.014%

curly braces and structure markers in code

code formatting and keywords

No Comments

No Known Activations