INDEX

Explanations

proper nouns and specific entities

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

↵

-0.46

↵↵

-0.38

-0.33

-0.31

-0.30

-0.27

ins

-0.27

-0.26

Ins

-0.26

ần

-0.26

POSITIVE LOGITS

 pleaſure

0.88

 itſelf

0.84

 queſta

0.82

 myſelf

0.78

ロウィン

0.78

 Geiſt

0.77

 queſto

0.77

 Monfieur

0.76

 fashiola

0.76

 faſt

0.75

Activations Density 1.421%

proper nouns and specific entities

No Comments

No Known Activations