INDEX

Explanations

terms related to governance and research methodologies

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

Word fragments that could be turned into nouns

np_acts-logits-general · gemini-2.0-flash

legal and news headers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

 sementes

-0.59

 Daß

-0.55

 mijne

-0.54

 quæ

-0.53

lamos

-0.53

gården

-0.52

 âme

-0.52

 stedet

-0.52

luß

-0.51

 nemlig

-0.51

POSITIVE LOGITS

 betweenstory

0.62

OF

0.61

ED

0.55

IsContent

0.55

ERS

0.54

ISTS

0.50

Ārējās

0.50

protoc

0.50

IES

0.49

ER

0.49

Activations Density 0.398%

terms related to governance and research methodologies

Word fragments that could be turned into nouns

legal and news headers

No Comments

No Known Activations