INDEX

Explanations

technical terms related to electrical characteristics and performance metrics

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

 TabLayout

-0.45

recated

-0.44

IComponent

-0.43

ceas

-0.43

нгред

-0.42

namics

-0.42

 GenerationType

-0.41

enido

-0.41

ðsíða

-0.41

uestion

-0.40

POSITIVE LOGITS

DIPSETTING

0.39

 quality

0.38

 Qualität

0.38

 rendah

0.35

 qualité

0.35

ocumented

0.34

 kyllä

0.34

󠁣

0.34

 Funktionalität

0.33

 Berliner

0.33

Activations Density 0.028%

technical terms related to electrical characteristics and performance metrics

No Comments

No Known Activations