INDEX

Explanations

sections and headings in documents

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

PMailer

-0.70

 informée

-0.63

يكب

-0.63

wiſe

-0.60

tanleria

-0.58

 odkazy

-0.57

 ब्रेकडाउन

-0.57

 esternos

-0.57

ſelves

-0.57

 الرياضيه

-0.56

POSITIVE LOGITS

GeneratedValue

0.35

 Charles

0.34

pag

0.33

 paid

0.33

 group

0.32

Cam

0.31

varargin

0.31

 Mare

0.30

 pinulongan

0.30

 Վերցված

0.30

Activations Density 0.001%

sections and headings in documents

No Comments

No Known Activations