INDEX

Explanations

codes or identifiers related to computing processes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rizona

-0.58

 Chwiliwch

-0.57

 nimic

-0.55

 appunto

-0.54

thmetic

-0.52

 juſ

-0.52

OGND

-0.52

 cherchés

-0.50

請繼續往下閱讀

-0.48

annica

-0.48

POSITIVE LOGITS

0.42

0.40

0.39

0.37

lès

0.35

0.34

 betweenstory

0.34

jpg

0.33

Activations Density 0.653%

codes or identifiers related to computing processes

No Comments

No Known Activations