INDEX

Explanations

phrases that indicate consequences or implications related to decisions and actions

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

and its

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

Kilder

-0.40

ínguez

-0.36

 experto

-0.36

MediaStore

-0.35

 mecánico

-0.35

 filings

-0.35

 beina

-0.34

 matéria

-0.34

 instructions

-0.34

 oír

-0.34

POSITIVE LOGITS

 Infórmanos

0.56

RegressionTest

0.54

styleType

0.48

 taxpayers

0.48

RTEX

0.47

__*/

0.47

RegistryLite

0.47

 humankind

0.47

 taxpayer

0.46

WriteBarrier

0.46

Activations Density 0.018%

phrases that indicate consequences or implications related to decisions and actions

and its

No Comments

No Known Activations