INDEX

Explanations

punctuation, specifically brackets and parentheses, often indicating contextual or temporal references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

żdy

-0.46

unque

-0.46

 gains

-0.44

有一天

-0.42

kwds

-0.41

nodeValue

-0.40

 bell

-0.40

 schie

-0.39

нятно

-0.39

one

-0.39

POSITIVE LOGITS

contentLoaded

0.95

withIdentifier

0.83

 وتسجيلات

0.82

 wireType

0.76

SuppressLint

0.73

 AssemblyCompany

0.72

 âgé

0.71

 AssemblyTitle

0.69

 يتيمه

0.69

 محفوظة

0.69

Activations Density 0.280%

punctuation, specifically brackets and parentheses, often indicating contextual or temporal references

No Comments

No Known Activations