INDEX

Explanations

instances of the word "too" and its variations indicating excessiveness or inclusion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 14-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_14/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.14.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

 hẳn

-0.35

<?

-0.33

rawtypes

-0.32

 encuentras

-0.31

 रो

-0.31

 encantó

-0.31

ウンサー

-0.30

 appena

-0.30

 Acción

-0.30

已

-0.30

POSITIVE LOGITS

ckså

0.71

自分も

0.63

 EClass

0.62

SharedCtor

0.59

こちらも

0.58

titleMargin

0.58

 HasFactory

0.58

 mergeFrom

0.56

 oprot

0.55

inSlope

0.55

Activations Density 0.300%

instances of the word "too" and its variations indicating excessiveness or inclusion

No Comments

No Known Activations