INDEX

Explanations

references to cultural or historical figures and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_23/width_16k/average_l0_75

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

 personne

-0.67

 miembro

-0.53

 miembros

-0.53

 member

-0.52

 membre

-0.52

 Personne

-0.51

 membro

-0.51

 osoba

-0.50

 Membre

-0.50

 membros

-0.49

POSITIVE LOGITS

 wise

0.72

posedge

0.66

 Wise

0.65

 Sheikh

0.64

 king

0.62

Wise

0.62

 prophet

0.61

 sage

0.60

ScopeManager

0.60

 Duke

0.58

Activations Density 0.392%

references to cultural or historical figures and titles

No Comments

No Known Activations