INDEX

Explanations

phrases that reference varieties, classifications, or groups within a broader context

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

はじめに

-0.54

KommentareTeilen

-0.43

 préfé

-0.40

parsedMessage

-0.39

makeText

-0.39

orologio

-0.37

preferred

-0.36

VersionUID

-0.36

 convenience

-0.35

TProtocol

-0.35

POSITIVE LOGITS

tanleria

0.47

 Chwiliwch

0.46

SequentialGroup

0.46

 '\\;'

0.40

 quite

0.40

":[{

0.39

addContainerGap

0.38

omock

0.38

ốc

0.38

 truly

0.37

Activations Density 0.117%

phrases that reference varieties, classifications, or groups within a broader context

No Comments

No Known Activations