INDEX

Explanations

terms related to authorship and editorial roles

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

ending in "ial", "ally", or similar suffixes

np_acts-logits-general · gemini-2.0-flash

ally or ial suffixes

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Bradbury

-0.50

lawayo

-0.47

 McIntyre

-0.47

baran

-0.46

 Baseline

-0.46

 claro

-0.46

 Glä

-0.46

ness

-0.46

 folle

-0.45

 Blanket

-0.45

POSITIVE LOGITS

ally

1.17

IAL

1.17

ial

1.11

ially

1.10

ials

1.04

inal

1.04

nal

0.96

yal

0.94

IALS

0.94

cial

0.94

Activations Density 0.364%

terms related to authorship and editorial roles

ending in "ial", "ally", or similar suffixes

ally or ial suffixes

No Comments

No Known Activations