INDEX

Explanations

references to locations and directions

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

transQ

-0.46

 חיצוניים

-0.39

LookAnd

-0.39

RegressionTest

-0.39

verwijspagina

-0.38

detectChanges

-0.38

findpost

-0.36

 productivo

-0.35

 <<<<<<<<<<<<<<

-0.33

Sucesor

-0.33

POSITIVE LOGITS

༘

0.51

CloseOperation

0.51

 fusca

0.48

 Waſſer

0.46

 lópez

0.45

 terminal

0.45

 stockbild

0.45

 Launch

0.44

Itoa

0.43

 Geiſt

0.43

Activations Density 0.005%

references to locations and directions

No Comments

No Known Activations