INDEX

Explanations

discourse markers that indicate a continuation or elaboration in thought

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

introducing a statement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

ic

-0.36

 procedimento

-0.33

 invokingState

-0.32

โย

-0.30

 actie

-0.29

 Terraria

-0.29

幸いです

-0.29

 çat

-0.28

 silenzio

-0.28

Judge

-0.28

POSITIVE LOGITS

DockStyle

0.60

irited

0.59

 Motta

0.57

 Pascual

0.56

 المعيارى

0.54

########.

0.54

 Palacios

0.53

ValueStyle

0.52

monary

0.50

الحياه

0.50

Activations Density 0.002%

discourse markers that indicate a continuation or elaboration in thought

introducing a statement

No Comments

No Known Activations