INDEX

Explanations

specific names, titles, and academic formatting related to citations or references

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

Ārējās

-0.44

 trânsito

-0.42

 femininas

-0.41

 aéreo

-0.41

 llorando

-0.40

 segredo

-0.40

 inspirasi

-0.39

 estacionamento

-0.39

 servire

-0.39

 łą

-0.39

POSITIVE LOGITS

----</

0.65

 kasarigan

0.61

rungsseite

0.56

BeginInit

0.50

Personendaten

0.50

 TextAlign

0.49

 ujednoznacz

0.46

stateProvider

0.46

 pref

0.45

ⓧ

0.43

Activations Density 0.036%

specific names, titles, and academic formatting related to citations or references

No Comments

No Known Activations