INDEX

Explanations

locations and geographical names

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

hyphenated names and locations

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

 řízení

-0.48

 Grüsse

-0.43

 técn

-0.43

ßerdem

-0.40

 sebagainya

-0.40

 éstos

-0.39

guisement

-0.39

 vuestros

-0.39

 שוליים

-0.38

 éstas

-0.38

POSITIVE LOGITS

0.88

®-

0.70



0.66

/*-

0.64

−

0.62

\-

0.62

‐

0.62

&-

0.61

﹣

0.61

–

0.60

Activations Density 0.035%

locations and geographical names

hyphenated names and locations

No Comments

No Known Activations