INDEX

Explanations

specific programming commands and keywords related to data processing and manipulation

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

train CNN

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

ⓧ

-0.38

 ſtate

-0.37

 ſur

-0.36

dBm

-0.34

 éve

-0.34

OGND

-0.33

 autorytatywna

-0.33

 ſche

-0.31

guenos

-0.31

%)$

-0.31

POSITIVE LOGITS

enderror

0.65

XmlAccessType

0.60

 otomatig

0.57

GraphicsUnit

0.57

endregion

0.56

.*")]

0.55

gye

0.50

 FetchType

0.48

vece

0.47

Vidite

0.47

Activations Density 0.141%

specific programming commands and keywords related to data processing and manipulation

train CNN

No Comments

No Known Activations