INDEX

Explanations

programming-related terms and constructs

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

property and field accessors

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

optionalTypeArgs

-0.60

RectangleBorder

-0.57

 ligiloj

-0.57

esModule

-0.52

ंदीखरीदारी

-0.52

PageRoute

-0.52

MLLoader

-0.52

 BrowserModule

-0.52

DockStyle

-0.51

ValueStyle

-0.51

POSITIVE LOGITS

InputTagHelper

0.51

FDRE

0.41

éndole

0.38

entliche

0.37

OwnProperty

0.37

 capilla

0.36

 gavetas

0.36

ี้ยง

0.35

Fitur

0.35

øb

0.35

Activations Density 0.053%

programming-related terms and constructs

property and field accessors

No Comments

No Known Activations