INDEX

Explanations

the verb "to be" used in various forms like "are" and "is."

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

what are or what is questions

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

 queſta

-0.89

posedge

-0.77

 betweenstory

-0.70

erdings

-0.68

存于互联网档案馆

-0.68

parsedMessage

-0.65

AnchorStyles

-0.62

SPJ

-0.60

 disambiguazione

-0.59

 nakalista

-0.59

POSITIVE LOGITS

 what

0.90

 What

0.80

What

0.79

what

0.79

 WHAT

0.68

whats

0.64

the

0.63

 whats

0.58

Whats

0.56

WHAT

0.56

Activations Density 0.040%

the verb "to be" used in various forms like "are" and "is."

what are or what is questions

No Comments

No Known Activations