INDEX

Explanations

URLs and domain-related terms within a technical context

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

external links

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

Personendaten

-0.77

 ModelExpression

-0.61

 zwiſchen

-0.57

 disambiguazione

-0.54

 ujednoznacz

-0.54

 nawr

-0.52

StructEnd

-0.51

didSet

-0.50

AddTagHelper

-0.50

IsContent

-0.50

POSITIVE LOGITS

Välislingid

0.39

webElementXpaths

0.33

 насељу

0.29

 Signalez

0.28

省市镇

0.28

inscrire

0.28

CDCl

0.28

getreten

0.28

发表于

0.27

DidLoad

0.27

Activations Density 0.000%

URLs and domain-related terms within a technical context

external links

No Comments

No Known Activations