INDEX

Explanations

mathematical notation related to limits and functions

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

<eos>

-0.61

 vôtre

-0.53

 betweenstory

-0.50

 nôtre

-0.49

 Against

-0.45

buttonBar

-0.45

 mémor

-0.43

ers

-0.43

ControllerTest

-0.42

against

-0.42

POSITIVE LOGITS

 AssemblyTitle

0.77

WriteAttribute

0.66

httphttps

0.64

enumii

0.62

ftagPool

0.61

mi

0.60

enumi

0.59

<i>

0.59

__(/*!

0.56

NameInMap

0.54

Activations Density 0.002%

mathematical notation related to limits and functions

No Comments

No Known Activations