INDEX

Explanations

references to attacks or hostile actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 թվական

-0.70

setOnItem

-0.65

antaranya

-0.61

nO

-0.61

Ав

-0.60

Roe

-0.60

ordano

-0.60

Roe

-0.60

lblNewLabel

-0.59

POSITIVE LOGITS

 Attack

2.57

 attack

2.56

 attacks

2.42

 ATTACK

2.38

attack

2.37

 Attacks

2.34

Attack

2.23

Attacks

2.21

ATTACK

2.20

attacks

2.15

Activations Density 0.043%

references to attacks or hostile actions

No Comments

No Known Activations