INDEX

Explanations

phrases indicating the rationale or justification for certain statements or actions

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

reason why

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

accumulative

-0.46

retto

-0.41

evid

-0.41

Yas

-0.40

 sart

-0.40

 Laure

-0.40

vstack

-0.39

 meanwhile

-0.38

ایج

-0.38

cotch

-0.38

POSITIVE LOGITS

 AssemblyCulture

0.57

 proč

0.50

 Wicidata

0.50

Билгалдахарш

0.46

:✨

0.45

 kenapa

0.44

 choosing

0.41

 telefónica

0.41

 preferências

0.40

esez

0.40

Activations Density 0.048%

phrases indicating the rationale or justification for certain statements or actions

reason why

No Comments

No Known Activations