INDEX

Explanations

references to specific characters or terms associated with a storyline

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

starting with "ot", "kot", "mot", or "bot"

np_acts-logits-general · gemini-2.0-flash

Kot, Ot, bot, Rot, motti

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

withIdentifier

-0.66

 pleaſure

-0.59

ście

-0.58

 tartalomajánló

-0.58

الإنجليزية

-0.56

AndEndTag

-0.56

 Chriftian

-0.54

 censiti

-0.54

IVEREF

-0.54

пания

-0.53

POSITIVE LOGITS

setSubject

0.53

Rot

0.53

jec

0.53

Fot

0.49

issex

0.48

SOT

0.48

Wot

0.47

arot

0.47

Bot

0.47

bouts

0.47

Activations Density 0.200%

references to specific characters or terms associated with a storyline

starting with "ot", "kot", "mot", or "bot"

Kot, Ot, bot, Rot, motti

No Comments

No Known Activations