INDEX

Explanations

references to honorific titles or terms of respect

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

Sucesor

-0.58

ValueStyle

-0.55

Spoljašnje

-0.55

 estekak

-0.55

stateProvider

-0.53

issippi

-0.50

nommen

-0.49

apunov

-0.48

 denn

-0.47

obby

-0.46

POSITIVE LOGITS

HON

0.88

HON

0.86

hon

0.77

 onor

0.77

 sauvages

0.75

 dishon

0.70

Hon

0.69

Hon

0.69

 olympique

0.68

hon

0.66

Activations Density 0.069%

references to honorific titles or terms of respect

No Comments

No Known Activations