INDEX

Explanations

references to specific places, organizations, or formal structures within a document

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 kasarigan

-0.87

 AssemblyCulture

-0.85

 CreateTagHelper

-0.79

 noDo

-0.75

__(/*!

-0.73

WriteTagHelper

-0.73

 informée

-0.73

 EconPapers

-0.71

verwijspagina

-0.70

 ब्रेकडाउन

-0.69

POSITIVE LOGITS

sibirsk

0.35

])/

0.30

 rela

0.30

 fonte

0.29

แข

0.29

cir

0.28

==============

0.27

'/',

0.27

 fuente

0.26

 ситуа

0.26

Activations Density 1.238%

references to specific places, organizations, or formal structures within a document

No Comments

No Known Activations