INDEX

Explanations

concepts related to trust and reliability in studies or systems

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

technical property name suffixes

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

Xna

-0.52

 initComponents

-0.51

 Plum

-0.49

 wireType

-0.49

énario

-0.49

Gön

-0.47

medriver

-0.47

verwijspagina

-0.46

rungsseite

-0.46

 Messenger

-0.45

POSITIVE LOGITS

NameInMap

0.52

 Nachhaltigkeit

0.42

 prywat

0.38

 katholischen

0.38

BagLayout

0.36

 culturelle

0.35

 qualités

0.35

 caractère

0.34

 Sünde

0.33

 científicas

0.33

Activations Density 0.090%

concepts related to trust and reliability in studies or systems

technical property name suffixes

No Comments

No Known Activations