INDEX

Explanations

terms associated with technology and its implications

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

qualities and capabilities

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

 שוליים

-0.54

 CreateTagHelper

-0.53

 snippetHide

-0.51

MockBean

-0.49

 AssemblyCulture

-0.48

jetunion

-0.47

UnknownFieldSet

-0.47

wahlen

-0.47

ConstraintMaker

-0.47

Personensuche

-0.46

POSITIVE LOGITS

 intptr

0.40

enegger

0.35

 simplicity

0.34

asable

0.34

mountable

0.33

 root

0.32

subpackage

0.32

itize

0.31

 scalable

0.31

ttemberg

0.31

Activations Density 0.026%

terms associated with technology and its implications

qualities and capabilities

No Comments

No Known Activations