INDEX

Explanations

nested function calls and their parameters

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 volverá

-0.42

 continuará

-0.39

ckså

-0.35

 coucher

-0.34

 continú

-0.32

farande

-0.32

 continúa

-0.32

 préfé

-0.32

styleUrls

-0.31

fährt

-0.31

POSITIVE LOGITS

 astore

0.94

 propOrder

0.71

ſicht

0.68

 kasarigan

0.68

 autorytatywna

0.66

 deſſen

0.64

<unused51>

0.62

<unused68>

0.62

<unused28>

0.62

<unused8>

0.62

Activations Density 0.110%

nested function calls and their parameters

No Comments

No Known Activations