INDEX

Explanations

terms related to secrecy and hidden concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

///</

-0.49

-0.48

etcode

-0.39

 useRouter

-0.39

Personensuche

-0.36

kháu

-0.36

 situasi

-0.35

situ

-0.35

oredCriteria

-0.35

 rapides

-0.35

POSITIVE LOGITS

 doInBackground

0.63

tagHelperRunner

0.57

 treasures

0.57

MethodManager

0.57

 disambiguazione

0.56

 BoxFit

0.54

 behind

0.54

 hide

0.53

 agendas

0.53

 "..\..\

0.52

Activations Density 0.340%

terms related to secrecy and hidden concepts

No Comments

No Known Activations