INDEX

Explanations

references to formatting and layout attributes in text or code

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

text padding and alignment

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

 '\\;'

-0.61

 ब्रेकडाउन

-0.55

 Administrativna

-0.53

tvguidetime

-0.53

KommentareTeilen

-0.52

 surla

-0.50

 дописавши

-0.49

 AssemblyCompany

-0.48

__(/*!

-0.48

RenderAtEndOf

-0.47

POSITIVE LOGITS

obec

0.37

 setw

0.36

setVerticalGroup

0.35

 vertical

0.35

stmt

0.33

 cáo

0.33

 clearColor

0.33

letten

0.32

енча

0.32

مراجع

0.32

Activations Density 0.098%

references to formatting and layout attributes in text or code

text padding and alignment

No Comments

No Known Activations