INDEX

Explanations

pronouns and modal verbs indicating ability or permission

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

predicates, Yoda, yielded, give

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

الحياه

-0.50

ReusableCell

-0.49

المناصب

-0.47

BagConstraints

-0.45

tableFuture

-0.45

 ſch

-0.43

ceps

-0.43

ChildScrollView

-0.42

ſelf

-0.41

 ſche

-0.41

POSITIVE LOGITS

makeConstraints

0.69

 otomatig

0.61

ⓧ

0.57

EDEFAULT

0.55

참고

0.50

windowFixed

0.48

oredCriteria

0.47

PYX

0.45

 ویکی‌پدی

0.44

evos

0.44

Activations Density 0.013%

pronouns and modal verbs indicating ability or permission

predicates, Yoda, yielded, give

No Comments

No Known Activations