INDEX

Explanations

negative contractions and related phrases indicating inability or denial

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

cannot or can't

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 nahilalakip

-0.72

الحياه

-0.65

 الحره

-0.64

 ProtoMessage

-0.64

 يتيمه

-0.63

enterOuterAlt

-0.62

 متعلقه

-0.59

richTextPanel

-0.58

 సౌకర్య

-0.55

+#+#

-0.54

POSITIVE LOGITS

 cannot

0.57

cannot

0.55

 Cannot

0.54

Cannot

0.53

 cant

0.47

 нельзя

0.47

can

0.43

 CANNOT

0.43

 不能

0.38

不可以

0.38

Activations Density 0.025%

negative contractions and related phrases indicating inability or denial

cannot or can't

No Comments

No Known Activations