INDEX

Explanations

tallies of significant events or achievements across various contexts

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

first [something]

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

 <>",

-0.44

 సౌకర్య

-0.40

awtextra

-0.39

 يتيمه

-0.35

Бахар

-0.35

 menarik

-0.35

 stadig

-0.34

 udaler

-0.33

ginfo

-0.33

 setia

-0.33

POSITIVE LOGITS

principalColumn

0.46

HostException

0.44

addCriterion

0.44

endix

0.42

 BoxFit

0.42

""}

0.42

 CommonModule

0.42

ḏ

0.41

%");

0.41

艋

0.41

Activations Density 0.032%

tallies of significant events or achievements across various contexts

first [something]

No Comments

No Known Activations