INDEX

Explanations

references to academic papers and citations

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

 casada

-0.33

 Störungen

-0.31

 Fernsehen

-0.31

 Einbau

-0.30

 porém

-0.30

 táctil

-0.30

 complejidad

-0.30

 Bescheid

-0.29

 Berechnung

-0.29

 Trocken

-0.29

POSITIVE LOGITS

XmlRootElement

0.79

contentLoaded

0.64

原始内容存档于

0.64

Personendaten

0.63

GEBURTS

0.61

VersionUID

0.60

canestro

0.59

TypedDataSet

0.56

日閲覧

0.55

msgTypes

0.54

Activations Density 0.415%

references to academic papers and citations

No Comments

No Known Activations