INDEX

Explanations

proper nouns and specific names

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

 however

-0.37

 However

-0.35

 derfor

-0.34

 following

-0.34

 fact

-0.32

However

-0.32

 journées

-0.32

 Beine

-0.31

↵↵

-0.30

She

-0.30

POSITIVE LOGITS

 autorytatywna

0.89

 CanadaChoose

0.89

 EconPapers

0.86

iſen

0.82

OGND

0.81

ſicht

0.80

adaptiveStyles

0.80

 Paglinawan

0.80

 चीज़ों

0.79

Tembelea

0.78

Activations Density 1.442%

proper nouns and specific names

No Comments

No Known Activations