INDEX

Explanations

certain numeric identifiers, particularly those associated with records or references

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

numbers and citations

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

 незавершена

-0.64

Personensuche

-0.57

 linkovi

-0.56

 ComVisible

-0.52

 ujednoznacz

-0.52

CPtr

-0.52

DrawerToggle

-0.51

CodeAttribute

-0.50



-0.50

 resourceCulture

-0.49

POSITIVE LOGITS

Odkazy

0.46

 partij

0.39

yet

0.38

 freundlichen

0.37

该公司

0.36

Kilder

0.35

 productivo

0.35

 interessieren

0.35

curities

0.35

 zainteres

0.34

Activations Density 0.000%

certain numeric identifiers, particularly those associated with records or references

numbers and citations

No Comments

No Known Activations