INDEX

Explanations

specific names and titles associated with individuals and their roles

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Paglinawan

-0.51

 estekak

-0.51

InSection

-0.50

RTCK

-0.49

MessageTagHelper

-0.49

exitRule

-0.48

 jsPsych

-0.47

CodeAttribute

-0.46

InstrumentedTest

-0.46

WindowConstants

-0.46

POSITIVE LOGITS

annica

0.53

prefixer

0.53

liferay

0.46

gine

0.43

aktur

0.43

xtext

0.42

maker

0.42

myn

0.42

Vtbl

0.41

 ListNode

0.40

Activations Density 0.076%

specific names and titles associated with individuals and their roles

No Comments

No Known Activations