INDEX

Explanations

references to individuals or parties involved in legal scenarios or cases

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

pronouns and conjunctions

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

rungsseite

-0.52

RegressionTest

-0.51

TestingModule

-0.51

脚注の使い方

-0.47

SuppressLint

-0.45

findpost

-0.44

Билгалдахарш

-0.44

NewUrlParser

-0.44

setContentView

-0.41

AndroidJUnit

-0.40

POSITIVE LOGITS

indd

0.41

 الحره

0.40

 насељу

0.40

niosek

0.40

yorsunuz

0.38

 вікі

0.37

ificato

0.37

anias

0.37

msleep

0.37

 للمعارف

0.36

Activations Density 0.001%

references to individuals or parties involved in legal scenarios or cases

pronouns and conjunctions

No Comments

No Known Activations