INDEX

Explanations

manufacturing and engineering-related terms

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

is followed by a verb or adjective

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

HomeAsUpEnabled

-0.30

unconfirmed

-0.30

onShow

-0.29

不如

-0.29

ʖ

-0.28

Greater

-0.28

 гри

-0.28

greater

-0.28

提

-0.27

timento

-0.27

POSITIVE LOGITS

MLLoader

0.70

 CreateTagHelper

0.66

الدراسه

0.65

usercontent

0.64

ⓧ

0.63

fillType

0.63

出版年

0.60



0.60

الحياه

0.59

AnchorStyles

0.58

Activations Density 0.104%

manufacturing and engineering-related terms

is followed by a verb or adjective

No Comments

No Known Activations