INDEX

Explanations

words or phrases related to identity, specifically non-identity terms and personal references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 40-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_40/width_16k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.40.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

 enfans

-0.49

 esprits

-0.47

SharedDtor

-0.39

 будешь

-0.39

XtraGrid

-0.37

 Continued

-0.37

 čierna

-0.36

Morning

-0.36

回事

-0.36

IBarButtonItem

-0.36

POSITIVE LOGITS

si

0.96

è

0.94

ha

0.90

era

0.69

ci

0.64

 viene

0.61

 può

0.60

fa

0.58

va

0.57

 deve

0.56

Activations Density 0.069%

words or phrases related to identity, specifically non-identity terms and personal references

No Comments

No Known Activations