INDEX

Explanations

phrases centered around value and benefit in various contexts

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

words following "for" or "to"

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

ClCompile

-0.45

iastes

-0.43

 defaultdict

-0.43

جوايز

-0.43

 lenker

-0.42

cupertino

-0.41

citep

-0.41

ⓧ

-0.40

NewUrlParser

-0.40

ɵɵelementStart

-0.39

POSITIVE LOGITS

tagHelperRunner

0.54

ukone

0.49

ValueStyle

0.49

Portale

0.48

 Infórmanos

0.48

 humankind

0.47

ChildScrollView

0.47

 mankind

0.46

 pensioners

0.44

 families

0.44

Activations Density 0.027%

phrases centered around value and benefit in various contexts

words following "for" or "to"

No Comments

No Known Activations