INDEX

Explanations

phrases that suggest making considered choices or recommendations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_11/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

RegistryLite

-0.53

 ModelExpression

-0.47

AndEndTag

-0.45

 يتيمه

-0.45

BeginInit

-0.44

ValueStyle

-0.43

 Савезне

-0.43

VolleyError

-0.42

Portail

-0.41

 NSCoder

-0.41

POSITIVE LOGITS

 advisable

0.70

 sebaiknya

0.65

 conviene

0.64

advised

0.64

 recomendable

0.62

recommended

0.61

 warto

0.60

 wisest

0.58

 recommended

0.57

 prudent

0.56

Activations Density 0.626%

phrases that suggest making considered choices or recommendations

No Comments

No Known Activations