INDEX

Explanations

references to specific individuals, roles, or identities within contexts, often linked to personal experiences or public statements

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

preceding professions or origin

np_acts-logits-general · gemini-2.0-flash

people's roles or identifiers

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

 pinulongan

-0.43

ftagPool

-0.40

 sèche

-0.38

Aholisi

-0.38

EndInit

-0.37

 húmedo

-0.37

 Vikipedi

-0.36

 mostaza

-0.36

 incluido

-0.35

 automático

-0.35

POSITIVE LOGITS

ⓧ

0.65

kuuta

0.49

Jährige

0.46

hynch

0.45

вушка

0.45

opsida

0.44

fertigt

0.44

 resourceCulture

0.44

 roadster

0.44

unier

0.43

Activations Density 0.143%

references to specific individuals, roles, or identities within contexts, often linked to personal experiences or public statements

preceding professions or origin

people's roles or identifiers

No Comments

No Known Activations