INDEX

Explanations

mathematical symbols and notation

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

specific mathematical or code contexts

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

RenderAtEndOf

-0.67

featureID

-0.62

UnusedPrivate

-0.60

parsedMessage

-0.59

 iNdEx

-0.59

ConstraintMaker

-0.53

ftagPool

-0.50

Personendaten

-0.50

UrlResolution

-0.49

PerformLayout

-0.47

POSITIVE LOGITS

ContentAlignment

0.36

Ответить

0.35

 outcome

0.35

aarrggbb

0.34

outcome

0.34

 juger

0.34

memset

0.33

casque

0.33

 spill

0.32

aufen

0.32

Activations Density 0.392%

mathematical symbols and notation

specific mathematical or code contexts

No Comments

No Known Activations