INDEX

Explanations

evidence and references to past actions or knowledge concerning legal or social situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

dataclass

-0.43

ArgsConstructor

-0.40

XtraBars

-0.38

 scribe

-0.35

 Singles

-0.35

stree

-0.34

neat

-0.33

Aktualisiert

-0.33

 saites

-0.33

 gawas

-0.32

POSITIVE LOGITS

 obvious

0.62

obvious

0.57

 évidence

0.56

 clear

0.52

 clearly

0.52

SharedDtor

0.51

 obvio

0.50

Obvious

0.50

 dezelve

0.48

testens

0.48

Activations Density 0.605%

evidence and references to past actions or knowledge concerning legal or social situations

No Comments

No Known Activations