INDEX

Explanations

locations and geographical references

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

places and their adjacent words

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

 araña

-0.40

zości

-0.38

ไตล

-0.37

 bỏng

-0.37

DockStyle

-0.37

 puentes

-0.36

 atribu

-0.36

heitsbild

-0.36

uș

-0.36

 mulut

-0.36

POSITIVE LOGITS

 الرياضيه

0.56

 onCancelled

0.56

𓃵

0.54

AsUp

0.52

httphttps

0.52

Personendaten

0.52

MemoryWarning

0.51

 فريبيس

0.50

andoah

0.48

 NavController

0.47

Activations Density 0.015%

locations and geographical references

places and their adjacent words

No Comments

No Known Activations