INDEX

Explanations

repeated instances of the word "do," indicating inquiries or actions proposed

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

Follows question words (would, do, did, can) asking for help

np_acts-logits-general · gemini-2.0-flash

modal verbs followed by pronouns

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ſelves

-0.88

MLLoader

-0.81

ſelf

-0.79

WithIOException

-0.79

bewerken

-0.77

 незавершена

-0.74

脚注の使い方

-0.74

 Houſe

-0.73

 Himo

-0.72

 houſe

-0.72

POSITIVE LOGITS

 they

1.22

you

1.19

we

1.15

he

1.06

it

0.98

the

0.97

0.79

 those

0.75

 these

0.75

0.73

Activations Density 0.130%

repeated instances of the word "do," indicating inquiries or actions proposed

Follows question words (would, do, did, can) asking for help

modal verbs followed by pronouns

No Comments

No Known Activations