INDEX

Explanations

references to document classes and formatting in a technical context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_7/width_16k/average_l0_92

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 betweenstory

-0.59

ficulties

-0.45

Географи

-0.45

 obstante

-0.44

ificantly

-0.43

ikistan

-0.42

withstanding

-0.42

Rgds

-0.42

 للاسماء

-0.42

theless

-0.41

POSITIVE LOGITS

awaiter

0.76

istoitu

0.66

 виправивши

0.62

ंदीखरीदारी

0.61

Pautan

0.59

 snippetHide

0.58

twimg

0.57

 intptr

0.55

❋

0.53

 noten

0.53

Activations Density 1.493%

references to document classes and formatting in a technical context

No Comments

No Known Activations