INDEX

Explanations

occurrences of the word "by."

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

Immediately preceded by "by"

np_acts-logits-general · gemini-2.0-flash

by followed by explanatory phrase

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

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 pleaſure

-0.57

leſs

-0.57

 uſed

-0.56

 ſal

-0.56

 føl

-0.56

 ſtate

-0.55

 feroit

-0.55

 ſtand

-0.55

 Jefus

-0.52

 itſelf

-0.52

POSITIVE LOGITS

by

1.22

By

1.15

by

1.14

BY

1.13

By

1.11

 getBy

1.03

BY

1.02

getBy

0.91

bys

0.90

selectBy

0.90

Activations Density 0.411%

occurrences of the word "by."

Immediately preceded by "by"

by followed by explanatory phrase

No Comments

No Known Activations