INDEX

Explanations

references to intense emotional responses, particularly shock and horror

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

amir

-0.49

เภ

-0.48

RegisterType

-0.48

nginx

-0.47

!*\

-0.46

 criterios

-0.45

 parseFrom

-0.45

◜

-0.44

 alberi

-0.44

beros

-0.44

POSITIVE LOGITS

 shock

4.77

 Shock

4.21

shock

4.18

Shock

3.95

 SHOCK

3.68

 shocks

3.60

 shocked

3.27

shocked

2.94

 shocking

2.63

hock

2.39

Activations Density 0.057%

references to intense emotional responses, particularly shock and horror

No Comments

No Known Activations