INDEX

Explanations

numerical values and sequences, particularly those related to measurements and statistical results

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

hexadecimal numbers and 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-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

存于互联网档案馆

-0.66

httphttps

-0.62

 للاسماء

-0.59

elemField

-0.54

IsContent

-0.50

CPtr

-0.50

Билгалдахарш

-0.49

الإنجليزية

-0.48

 հղումներ

-0.47

CodeAttribute

-0.46

POSITIVE LOGITS

RegressionTest

0.52

 Dirig

0.40

⊤

0.38

UnusedPrivate

0.38

Hallo

0.37

openha

0.36

tymologie

0.36

Tottenham

0.36

 Paglinawan

0.35

ResponseData

0.35

Activations Density 0.153%

numerical values and sequences, particularly those related to measurements and statistical results

hexadecimal numbers and identifiers

No Comments

No Known Activations