INDEX

Explanations

references to media, programming, and system processes in technical contexts

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

tokens followed by instructions

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

In

-0.41

to

-0.38

if

-0.38

To

-0.36

If

-0.36

as

-0.35

is

-0.35

He

-0.35

POSITIVE LOGITS

 تضيفلها

1.06

 disambiguazione

0.98

Diwedd

0.94

LookAnd

0.88

 ligiloj

0.81

ésultats

0.79

UseVisualStyle

0.79

Xna

0.77

Autoritní

0.77

AnimationsModule

0.77

Activations Density 0.015%

references to media, programming, and system processes in technical contexts

tokens followed by instructions

No Comments

No Known Activations