INDEX

Explanations

references to absolute values or measurements in a scientific context

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

absolute plus specific noun

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

 biais

-0.45

Movies

-0.43

 cartera

-0.43

 Pests

-0.43

oeuvre

-0.42

Mereka

-0.42

MyApp

-0.40

Schriften

-0.40

 Wine

-0.39

 Smell

-0.39

POSITIVE LOGITS

 absolute

1.16

 Absolute

1.09

Absolute

1.05

absolute

1.04

 absolu

0.96

 ABSOL

0.90

 assolu

0.87

 absolut

0.82

ABSOL

0.81

 absol

0.80

Activations Density 0.007%

references to absolute values or measurements in a scientific context

absolute plus specific noun

No Comments

No Known Activations