INDEX

Explanations

terms and concepts related to deception and falsification, particularly in the context of news and information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

Revenir

-0.60

reactivex

-0.55

 pantaloni

-0.53

{{$

-0.53

 participaron

-0.52

 Graz

-0.51

bringung

-0.51

 pribadi

-0.50

--){

-0.50

 picioare

-0.50

POSITIVE LOGITS

 Fake

0.78

 fake

0.74

RTLD

0.73

Faux

0.73

Fake

0.71

LookAnd

0.71

fake

0.68

 faux

0.68

mockito

0.67

 Mock

0.66

Activations Density 0.276%

terms and concepts related to deception and falsification, particularly in the context of news and information

No Comments

No Known Activations