INDEX

Explanations

numerical identifiers or codes related to specific entities or references

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

start of turn user

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

 cozin

-0.34

 cielos

-0.33

LayoutStyle

-0.31

まだ

-0.30

culada

-0.29

 livre

-0.29

ниципа

-0.29

 entweder

-0.29

')):

-0.29

()):

-0.29

POSITIVE LOGITS

 незавершена

0.62

 IBOutlet

0.56

 informée

0.55

 ब्रेकडाउन

0.55

 AssemblyTitle

0.54

 surla

0.53

verifyException

0.53

WriteTagHelper

0.53

complexContent

0.52

GEBURTSDATUM

0.51

Activations Density 0.019%

numerical identifiers or codes related to specific entities or references

start of turn user

No Comments

No Known Activations