INDEX

Explanations

statements that affirm or emphasize a point in conversation

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

Beginning of a response or statement

np_acts-logits-general · gemini-2.0-flash

commencement of utterance

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

 Autorisations

-0.49

newArrayList

-0.47

 ویکی‌پدی

-0.45

mistak

-0.44

зулта

-0.44

Dominik

-0.43

 linkovi

-0.43

IVEREF

-0.43

 defaultstate

-0.43

OGND

-0.42

POSITIVE LOGITS

Well

0.60

 Well

0.56

 yeah

0.52

嘛

0.51

Welp

0.50

Anyway

0.49

 Anyway

0.49

 Yeah

0.48

畢竟

0.48

 maybe

0.47

Activations Density 0.237%

statements that affirm or emphasize a point in conversation

Beginning of a response or statement

commencement of utterance

No Comments

No Known Activations