INDEX

Explanations

repeated references to measurements in a text

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

specific measurements

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

 kysy

-0.65

 piú

-0.60

 nahilalakip

-0.56

 aikaa

-0.56

 dreapta

-0.53

 dieux

-0.52

 Мексичка

-0.51

 šķ

-0.51

 elétrico

-0.50

 prohibido

-0.50

POSITIVE LOGITS

 measurements

1.85

 Measurements

1.75

measurements

1.67

 measurement

1.63

Measurements

1.61

 Measurement

1.56

measurement

1.55

Measurement

1.45

 MEASUREMENTS

1.38

 MEASUREMENT

1.28

Activations Density 0.011%

repeated references to measurements in a text

specific measurements

No Comments

No Known Activations