INDEX

Explanations

references to hierarchical rankings or positions marked as "top."

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

"top" followed by another word

np_acts-logits-general · gemini-2.0-flash

top of lists or positions

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

orianCalendar

-0.58

 pelúcia

-0.54

quelize

-0.53

 eventId

-0.50

 MotionEvent

-0.50

دانشنامهٔ

-0.50

uilla

-0.49

modelBuilder

-0.48

 Eilish

-0.47

 hernia

-0.47

POSITIVE LOGITS

top

1.32

Top

1.27

top

1.27

Top

1.23

TOP

1.13

TOP

1.09

getTop

1.02

ToTop

0.96

 Tops

0.95

 tops

0.94

Activations Density 0.105%

references to hierarchical rankings or positions marked as "top."

"top" followed by another word

top of lists or positions

No Comments

No Known Activations