INDEX

Explanations

punctuation marks and colons that imply lists or explanations

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 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

!&

-0.66

awtextra

-0.64

 nahilalakip

-0.62

ſelves

-0.61

ThroughAttribute

-0.60

Читати

-0.60

TagMode

-0.58

*&

-0.58

 Theſe

-0.57

 disambiguazione

-0.56

POSITIVE LOGITS

<blockquote>

2.53

":

0.64

':

0.58

):

0.57

0.56

</blockquote>

0.55

 quote

0.55

":

0.53

 QUOTE

0.53

rungsseite

0.53

Activations Density 0.108%

punctuation marks and colons that imply lists or explanations

No Comments

No Known Activations