INDEX

Explanations

occurrences of specific identifiers or tokens related to digital resources

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

ValueStyle

-0.59

 AssemblyTitle

-0.57

ViewFeatures

-0.46

DockStyle

-0.44

MigrationBuilder

-0.43

OGND

-0.43

 informée

-0.43

TemporalType

-0.42

 Wikimedijinoj

-0.42

AddTagHelper

-0.40

POSITIVE LOGITS

transQ

0.49

ttemberg

0.46

XmlAccessType

0.44

 Himo

0.44

Jura

0.42

tanleria

0.40

 importado

0.40

 betweenstory

0.40

锈钢

0.40

prüche

0.40

Activations Density 0.309%

occurrences of specific identifiers or tokens related to digital resources

No Comments

No Known Activations