INDEX

Explanations

terms related to nouns and noun phrases

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

httphttps

-0.61

தே

-0.56

Emilio

-0.54

oppa

-0.53

tagHelperRunner

-0.52

ilight

-0.52

ilosop

-0.52

addCriterion

-0.51

cipline

-0.50

 Pergamon

-0.50

POSITIVE LOGITS

 nouns

0.94

 noun

0.90

Noun

0.79

noun

0.79

 Noun

0.79

 productName

0.44

名詞

0.44

 name

0.43

 itemName

0.42

mouseup

0.42

Activations Density 0.008%

terms related to nouns and noun phrases

No Comments

No Known Activations