INDEX

Explanations

keywords related to software functionality and user instructions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

这意味着

-0.31

 ello

-0.31

 inconnu

-0.30

 ilha

-0.30

 éstos

-0.30

 éstas

-0.29

 aldea

-0.29

nungszeiten

-0.29

 primarily

-0.29

Judge

-0.29

POSITIVE LOGITS

 المعيارى

0.75

 defaultstate

0.69

ंदीखरीदारी

0.65

ArgsConstructor

0.62

NameInMap

0.60

WriteTagHelper

0.60

DockStyle

0.59

Tembelea

0.59

\{\\

0.59

 насељу

0.57

Activations Density 0.245%

keywords related to software functionality and user instructions

No Comments

No Known Activations