INDEX

Explanations

phrases that question the credibility or reality of assertions

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

DockStyle

-0.70

IntoConstraints

-0.53

ⓧ

-0.50

 disambiguazione

-0.48

fjspx

-0.48

فحة

-0.47

rather

-0.45

 interesan

-0.45

 resourceCulture

-0.44

Rather

-0.41

POSITIVE LOGITS

 supposed

0.46

 supposedly

0.46

 excuse

0.42

 pretext

0.42

 excuses

0.42

 fitness

0.40

 pretended

0.40

 Tycoon

0.40

 serious

0.40

 яко

0.39

Activations Density 0.758%

phrases that question the credibility or reality of assertions

No Comments

No Known Activations