INDEX

Explanations

references to notable events or entities in a competitive context

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

GraphicsUnit

-0.81

EndContext

-0.63

 AssemblyProduct

-0.57

 geldig

-0.54

ParallelGroup

-0.54

 оригіналу

-0.54

 IndexPath

-0.48

 EconPapers

-0.47

Становништво

-0.46

UseVisualStyle

-0.46

POSITIVE LOGITS

GrantedAuthority

0.64

ftagPool

0.62

initializeApp

0.61

 հղումներ

0.59

 مشين

0.59

 وتسجيلات

0.58

0.57

 виправивши

0.57

 BioLib

0.55

/\.

0.55

Activations Density 0.297%

references to notable events or entities in a competitive context

No Comments

No Known Activations