INDEX

Explanations

instructions or procedural steps related to tasks and maintenance

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rungsseite

-0.68

homonymie

-0.42

 artists

-0.40

 creators

-0.38

 programmers

-0.35

AnchorTagHelper

-0.35

 organisations

-0.34

partements

-0.33

WebServlet

-0.33

 artist

-0.33

POSITIVE LOGITS

torque

0.53

IntoConstraints

0.52

 мәкал

0.51

 kaarangay

0.51

featureID

0.51

 Torque

0.50

 bolts

0.50

inWeight

0.49

 ſte

0.48

Torque

0.48

Activations Density 0.007%

instructions or procedural steps related to tasks and maintenance

No Comments

No Known Activations