INDEX

Explanations

political and electoral references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

Cau

-0.37

lier

-0.32

ско

-0.31

Sah

-0.30

 Gret

-0.30

Cho

-0.29

Str

-0.29

ServletContext

-0.29

 finally

-0.29

ực

-0.29

POSITIVE LOGITS

 betweenstory

0.69

UnsafeEnabled

0.69

 reelection

0.68

تقاوى

0.65

IsMutable

0.64

ագրություններ

0.64

RegressionTest

0.63

 propOrder

0.61

 renewal

0.60

 архивлан

0.57

Activations Density 0.276%

political and electoral references

No Comments

No Known Activations