INDEX

Explanations

references to numerical and methodological evaluations in research

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

ftagPool

-0.38

 ihop

-0.34

 Мексичка

-0.34

:][

-0.34

 liberal

-0.33

 Liberal

-0.32

SupportActionBar

-0.31

 cares

-0.31

 iParam

-0.31

⍢

-0.30

POSITIVE LOGITS

 invokingState

0.56

fixes

0.53

 snippetHide

0.52

PointerException

0.49

MessageTagHelper

0.49

 Seeder

0.47

 NSCoder

0.47

 Versus

0.47

༘

0.46

afficheront

0.45

Activations Density 0.042%

references to numerical and methodological evaluations in research

No Comments

No Known Activations