INDEX

Explanations

references to session data and user-context variables

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

configuration or code keywords

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

 المعيارى

-0.79

AndEndTag

-0.75

AddTagHelper

-0.73

Autoritní

-0.70

 Chwiliwch

-0.66

complexContent

-0.65

niſſe

-0.64

rungsseite

-0.63

 الحره

-0.62

 unſer

-0.62

POSITIVE LOGITS

sidemargin

0.40

 betweenstory

0.39

 espoir

0.35

 sidang

0.35

 navideñas

0.31

bcryptjs

0.30

 keja

0.30

 hiver

0.30

 juridiques

0.30

 musicales

0.30

Activations Density 0.009%

references to session data and user-context variables

configuration or code keywords

No Comments

No Known Activations