INDEX

Explanations

affirmative and negative phrases regarding actions and states

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

 Савезне

-0.79

IntoConstraints

-0.71

 betweenstory

-0.69

 >=",

-0.69

GraphicsUnit

-0.69

 pageContext

-0.68

tagez

-0.68

chitis

-0.67

 étions

-0.67

 SwitchCompat

-0.66

POSITIVE LOGITS

has

0.69

has

0.65

Has

0.60

Has

0.54

 Have

0.54

HAS

0.52

HAS

0.50

↵

0.49

 have

0.49

ve

0.47

Activations Density 0.133%

affirmative and negative phrases regarding actions and states

No Comments

No Known Activations