INDEX

Explanations

phrases indicating personal preferences or evaluations

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

Rüyada

-0.82

ConstraintMaker

-0.80

 ſind

-0.74

 дописавши

-0.72

 Exacts

-0.71

$_(

-0.71

 queſta

-0.70



-0.69

IsMutable

-0.68

 Anſ

-0.67

POSITIVE LOGITS

0.34

rechnung

0.34

 показать

0.32

 concentration

0.32

pit

0.32

ko

0.32

 Schme

0.32

LON

0.31

 stab

0.31

des

0.31

Activations Density 0.517%

phrases indicating personal preferences or evaluations

No Comments

No Known Activations