INDEX

Explanations

technical terms and parameters related to data retrieval and processing in code

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

 safety

-0.38

 custom

-0.37

ργο

-0.37

 Merlot

-0.35

 bell

-0.35

inkle

-0.34

HOB

-0.34

Hound

-0.34

 ProtoMessage

-0.33

 Louvain

-0.33

POSITIVE LOGITS

 noDo

0.60

 surla

0.60

AISSEE

0.51

 EconPapers

0.47

 ویکی‌پدیای

0.46

writeFieldEnd

0.45

 betweenstory

0.44

AddTagHelper

0.44

 дописавши

0.42

 виправивши

0.42

Activations Density 0.009%

technical terms and parameters related to data retrieval and processing in code

No Comments

No Known Activations