INDEX

Explanations

terms related to definitions and constructions in mathematical contexts

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



-0.71

 disambiguazione

-0.68

RuleContext

-0.65

 يتيمه

-0.65

 NSCoder

-0.64

interopRequire

-0.63

 ModelRenderer

-0.61

 transfieras

-0.61

 Paglinawan

-0.60

 lenker

-0.59

POSITIVE LOGITS

 definition

0.37

Havolalar

0.35

definition

0.33

vac

0.32

 recal

0.32

 définition

0.32

HAS

0.31

ful

0.31

ælp

0.29

 exactly

0.29

Activations Density 0.251%

terms related to definitions and constructions in mathematical contexts

No Comments

No Known Activations