INDEX

Explanations

numerical values associated with measurements or references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_23/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

Datuak

-0.51

 חיצוניים

-0.41

DidLoad

-0.41

 utafitiHapana

-0.41

 Ruhe

-0.40

ɵɵ

-0.40

FormTagHelper

-0.38

 Winaray

-0.37

Bibliograf

-0.36

страницу

-0.36

POSITIVE LOGITS

AndEndTag

0.61

RenderAtEndOf

0.49

+#+#

0.49

 فريبيس

0.48

pyx

0.48

UnusedPrivate

0.47

unknownFields

0.47

 CanadaChoose

0.45

 grains

0.43

 '\\;'

0.43

Activations Density 0.139%

numerical values associated with measurements or references

No Comments

No Known Activations