INDEX

Explanations

references to multiple parts or editions of a narrative or document

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

volume, edition, part, or installment numbers

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

CROSSTALK

-0.39

 nakalista

-0.38

 Infórmanos

-0.37

 hendes

-0.36

AndEndTag

-0.35

SourceChecksum

-0.33

 Trouver

-0.32

帖最后由

-0.32

SND

-0.31

 विश्वसनीयता

-0.31

POSITIVE LOGITS

bewerken

0.55

ніципалі

0.52

 gepubliceerd

0.52

 MutableLiveData

0.47

 Geiſt

0.47

styleType

0.46

SharedDtor

0.46

 Biôgrafia

0.46

chartInstance

0.46

RUnlock

0.46

Activations Density 0.250%

references to multiple parts or editions of a narrative or document

volume, edition, part, or installment numbers

No Comments

No Known Activations