INDEX

Explanations

punctuation marks and formatting symbols within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 يتيمه

-0.89

 queſta

-0.63

帖最后由

-0.59

⤹

-0.59

writeField

-0.58

niſſe

-0.57

舺

-0.54

 ſind

-0.54

antası

-0.53

новништво

-0.52

POSITIVE LOGITS

 prêtre

0.42

GOTREF

0.42

ømme

0.39

 fédé

0.38

 Later

0.38

 referrerpolicy

0.36

 Tags

0.35

 stampa

0.34

ResumeLayout

0.34

 phẩm

0.34

Activations Density 0.052%

punctuation marks and formatting symbols within the text

No Comments

No Known Activations