INDEX

Explanations

terms related to capabilities and measurements in research contexts

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

measurement of abstract qualities

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

 république

-0.40

UrlResolution

-0.38

 manhã

-0.37

findpost

-0.37

 camiset

-0.36

 alfombra

-0.35

 transfieras

-0.35

 algodón

-0.35

AutoField

-0.34

 gynhyrchwyd

-0.34

POSITIVE LOGITS

 تضيفلها

0.54

 assessment

0.53

edance

0.52

 soba

0.51

tagHelperRunner

0.51

opress

0.50

ccini

0.49

 analysis

0.49

TacToe

0.49

nerability

0.49

Activations Density 0.128%

terms related to capabilities and measurements in research contexts

measurement of abstract qualities

No Comments

No Known Activations