INDEX

Explanations

conjunctions or phrases that connect ideas

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

 testing

-0.48

RenderAtEndOf

-0.47

ﷻ

-0.47

enderror

-0.46

 initView

-0.44

FFS

-0.44

 pinulongan

-0.43

 Crew

-0.43

WriteLiteral

-0.42

 rasa

-0.42

POSITIVE LOGITS

 dumne

0.53

ailleurs

0.45

bitField

0.45

nobody

0.44

 mikä

0.43

 Daß

0.41

 nobody

0.40

 vocês

0.40

 luckily

0.40

แต่

0.39

Activations Density 0.200%

conjunctions or phrases that connect ideas

No Comments

No Known Activations