INDEX

Explanations

instances of numerical values and identifiers

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

parsedMessage

-0.98

 betweenstory

-0.98

 CreateTagHelper

-0.97

AndEndTag

-0.91

OGND

-0.90

 propOrder

-0.88

LookAnd

-0.86

 ويكيپ

-0.82

niſſe

-0.82

POSITIVE LOGITS

 nocturno

0.34

 begge

0.32

 Económica

0.32

 sanitaires

0.31

 gratuits

0.31

 Richtung

0.31

 peligros

0.29

 donnés

0.29

 inoxidable

0.28

 procédures

0.28

Activations Density 0.005%

instances of numerical values and identifiers

No Comments

No Known Activations