INDEX

Explanations

repeated uses of the word "re."

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

we're

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

/**

-0.66

 صوتيه

-0.63

uxxxx

-0.60

sns

-0.58

InjectAttribute

-0.53

 CascadeType

-0.52

pula

-0.52

 Acess

-0.50

 acess

-0.49

<?

-0.49

POSITIVE LOGITS

re

0.77

 were

0.69

 theyre

0.68

 Were

0.64

were

0.61

Were

0.60

 meille

0.60

are

0.59

 Estamos

0.56

 estamos

0.56

Activations Density 0.032%

repeated uses of the word "re."

we're

No Comments

No Known Activations