INDEX

Explanations

phrases or terms indicating questions and discussions about claims, evidence, or validity

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

question or statement completion

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

UnknownFieldSet

-0.60

MLLoader

-0.54

EndInit

-0.53

 eventId

-0.53

IContainer

-0.53

PROMOTED

-0.51

 ſind

-0.50

 leçons

-0.49

awtextra

-0.49

########.

-0.48

POSITIVE LOGITS

uxxxx

0.44

RTDA

0.41

헌

0.38

 estekak

0.37

IZO

0.36

seits

0.35

try

0.34

تقاوى

0.33

 ویکی‌پدی

0.33

ellido

0.32

Activations Density 0.096%

phrases or terms indicating questions and discussions about claims, evidence, or validity

question or statement completion

No Comments

No Known Activations