INDEX

Explanations

statements related to official announcements and reports

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

confirmed, announced, approved, warned

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

 iniziale

-0.43

 jspb

-0.42

};*/

-0.40

]=>

-0.40

 crescita

-0.40

 earlier

-0.40

 tahun

-0.39

 póź

-0.38

Див

-0.38

 kés

-0.38

POSITIVE LOGITS

 snippetHide

0.55

ніципа

0.49

 kasarigan

0.49

addContainerGap

0.47

signable

0.46

enschappelijke

0.45

AutoField

0.43

 arraigned

0.42

 Winfield

0.42

LEncoder

0.41

Activations Density 0.020%

statements related to official announcements and reports

confirmed, announced, approved, warned

No Comments

No Known Activations