INDEX

Explanations

references to future events or scheduled occurrences

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

time and dates

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

InputBorder

-0.53

 Sanderson

-0.48

évaluateur

-0.48

 beetje

-0.47

 Exacts

-0.46

PLIES

-0.41

 webdriver

-0.40

trische

-0.40

 Secund

-0.39

 ActionResult

-0.39

POSITIVE LOGITS

 estekak

0.54

 chi̍t

0.53

 محفوظة

0.46

 فريبيس

0.46

HomeAsUpEnabled

0.45

 tartalomajánló

0.44

 MainAxisSize

0.44

onedDateTime

0.42

المناصب

0.42

 zamanda

0.41

Activations Density 0.004%

references to future events or scheduled occurrences

time and dates

No Comments

No Known Activations