INDEX

Explanations

occurrences of hyperlinks or web addresses

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

citations and URLs

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

 CURIAM

-0.53

balleur

-0.50

WebElementEntity

-0.49

tvguidetime

-0.45

intios

-0.45

ніципа

-0.44

SharedDtor

-0.42

 kasarigan

-0.42

 aislado

-0.41

 мәкал

-0.40

POSITIVE LOGITS

fjspx

0.53

ьаж

0.45

AndEndTag

0.43

 Dummies

0.42

 ditto

0.42

ValueStyle

0.42

setcounter

0.41

the

0.40

 handleMessage

0.40

Wikipedia

0.40

Activations Density 0.026%

occurrences of hyperlinks or web addresses

citations and URLs

No Comments

No Known Activations