INDEX

Explanations

instructions or warnings related to file modifications

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

do not edit or modify

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

IsContent

-0.62

ImageContext

-0.61

KommentareTeilen

-0.52

 jspb

-0.50

 ProtoMessage

-0.47

InSection

-0.45

sizeCache

-0.43

ioutil

-0.41

Aiheesta

-0.40

Ārējās

-0.39

POSITIVE LOGITS

ükemmel

0.41

LabelTagHelper

0.41

合法

0.40

 PRESERV

0.40

 preserving

0.40

 estrictamente

0.39

qvarna

0.39

 Kuba

0.38

kháu

0.38

 legales

0.38

Activations Density 0.015%

instructions or warnings related to file modifications

do not edit or modify

No Comments

No Known Activations