INDEX

Explanations

quantitative data expressed in numerical formats

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_11/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 pinulongan

-0.73

httphttps

-0.70

AddTagHelper

-0.67

adaptiveStyles

-0.56

 defaultstate

-0.56

RTEX

-0.56



-0.54

msgTypes

-0.54

Попис

-0.53

ijiet

-0.53

POSITIVE LOGITS

IR

0.38

Far

0.38

uyez

0.34

Ar

0.32

Ir

0.32

Wi

0.32

errHandler

0.31

wi

0.31

ga

0.30

Ga

0.30

Activations Density 1.194%

quantitative data expressed in numerical formats

No Known Activations