INDEX

Explanations

phrases that indicate recognition or reputation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 6-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_6/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.6.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

dersfield

-0.69

AccessorTable

-0.67

 MonoBehaviour

-0.65

sprozess

-0.64

Mif

-0.61

adecimal

-0.60

 Madsen

-0.60

Derbyniad

-0.60

 erklären

-0.60

 للمعارف

-0.59

POSITIVE LOGITS

 znám

0.65

 known

0.61

 kjent

0.61

 reputation

0.58

 consistently

0.57

 destacan

0.56

ésia

0.55

 känd

0.54

 Bess

0.54

 characterized

0.54

Activations Density 0.133%

phrases that indicate recognition or reputation

No Comments

No Known Activations