INDEX

Explanations

phrases indicating specific measurements, statistics, or options related to various topics

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 المعيارى

-0.54

 CanadaChoose

-0.51

arrings

-0.50

StateToProps

-0.49

 useAppContext

-0.48

 propOrder

-0.48

 ſel

-0.47

 Poin

-0.47

⅘

-0.47

actéristi

-0.46

POSITIVE LOGITS

intios

0.40

 resourceCulture

0.38

willReturn

0.35

TagMode

0.35

 Bewußt

0.34

 pungkas

0.34

rmtree

0.32

pas

0.32

 zápas

0.32

RTEE

0.32

Activations Density 0.114%

phrases indicating specific measurements, statistics, or options related to various topics

No Comments

No Known Activations