INDEX

Explanations

references to specific technical components and models related to data representation and processing

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

UTF-8–encoded punctuation, especially smart/curly double quotation marks and their byte-sequence tokens.

oai_token-act-pair · gpt-5 Triggered by @arielgoldsteinlab

0

np_max-act-logits · gpt-5 Triggered by @arielgoldsteinlab

Explanation could not be parsed.

eleuther_acts_top20 · gpt-5 Triggered by @arielgoldsteinlab

0

np_max-act · gpt-5 Triggered by @arielgoldsteinlab

UTF-8 byte sequences for special characters

oai_token-act-pair · deepseek-r1 Triggered by @arielgoldsteinlab

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

RectangleBorder

-0.42

cstdio

-0.36

PropertyGroup

-0.35

/*

-0.34

 bağlantılar

-0.33

XmlAccessorType

-0.32

Slf

-0.32

][$

-0.32

伝

-0.30

{[

-0.30

POSITIVE LOGITS

fortawesome

0.65

 ویکی‌پدی

0.59

Jeografia

0.54

colhead

0.52

:✨

0.50

verifyException

0.50

Vidite

0.49

 Saras

0.48

ніципа

0.48

 Sault

0.48

Activations Density 1.096%

references to specific technical components and models related to data representation and processing

UTF-8–encoded punctuation, especially smart/curly double quotation marks and their byte-sequence tokens.

0

Explanation could not be parsed.

0

UTF-8 byte sequences for special characters

No Comments

No Known Activations