INDEX

Explanations

phrases related to lessons or recommendations in various contexts

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

Außerdem

-0.46

kì

-0.45

нибудь

-0.45

 zufolge

-0.39

:]

-0.37

 Außerdem

-0.34

PointerException

-0.34

 malheur

-0.32

 riguardo

-0.31

 <!--

-0.30

POSITIVE LOGITS

 propOrder

0.84

endipity

0.79

Folks

0.79

 impacting

0.76

Challenges

0.75

 leveraging

0.74

challenged

0.74

 impacted

0.73

 impactful

0.72

 utford

0.72

Activations Density 6.308%

phrases related to lessons or recommendations in various contexts

No Comments

No Known Activations