INDEX

Explanations

HTML metadata tags

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

"meta" or "beta" prefixes

np_acts-logits-general · gemini-2.0-flash

meta and metadata

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

 jongen

-0.44

unction

-0.41

 iyong

-0.37

Complaint

-0.35

 bucal

-0.35

 ejus

-0.35

prepareStatement

-0.34

Draw

-0.34

our

-0.34

 exclusiva

-0.34

POSITIVE LOGITS

meta

1.21

Meta

1.20

 meta

1.20

 Meta

1.18

 META

1.10

 méta

0.93

 metas

0.90

メタ

0.88

 metadata

0.87

META

0.87

Activations Density 0.285%

HTML metadata tags

"meta" or "beta" prefixes

meta and metadata

No Comments

No Known Activations