INDEX

Explanations

instances of user input and related prompts

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

IsContent

-0.81

ValueStyle

-0.75

berdayakan

-0.74

ChildScrollView

-0.72

 '\\;'

-0.68

EndContext

-0.63

UserScript

-0.62

WaitGroup

-0.60

AddTagHelper

-0.59

niſſe

-0.58

POSITIVE LOGITS

 antig

0.30

⋱

0.30

nEnter

0.29

OfYear

0.29

 Carlo

0.28

حياته

0.27

hum

0.27

Pop

0.27

Digite

0.27

lan

0.27

Activations Density 0.050%

instances of user input and related prompts

No Comments

No Known Activations