INDEX

Explanations

numerical values and references in the text

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

 ब्रेकडाउन

-1.04

 snippetHide

-1.02

IntoConstraints

-1.01

IsMutable

-1.00

 للمعارف

-0.99

 queſta

-0.99

ftagPool

-0.96

tagext

-0.95

UnusedPrivate

-0.93

 surla

-0.92

POSITIVE LOGITS

 jagung

0.29

December

0.28

 Münch

0.27

0.26

 Formación

0.25

 Andalucía

0.24

 Infantil

0.24

 Erscheinung

0.24

 nettsted

0.24

 Dabei

0.24

Activations Density 0.003%

numerical values and references in the text

No Comments

No Known Activations