INDEX

Explanations

phrases that express contradiction or nuanced emotions regarding experiences or 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

Портали

-0.38

 disambiguazione

-0.37

rikes

-0.36

 cooper

-0.34

enterOuterAlt

-0.34

DIPSETTING

-0.33

Zunanje

-0.33

 SimpleName

-0.32

Karakteristik

-0.32

 prepar

-0.32

POSITIVE LOGITS

iastes

0.48

styleType

0.47

 warts

0.47

 TextInputType

0.46

 imperfections

0.46

 imperfection

0.44

 imperfect

0.44

ToBounds

0.43

]")]

0.43

ThroughAttribute

0.43

Activations Density 0.485%

phrases that express contradiction or nuanced emotions regarding experiences or situations

No Comments

No Known Activations