INDEX

Explanations

references to requests and formal procedures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_22/width_16k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.hook_mlp_out

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

 EconPapers

-0.99

cshtml

-0.65

 ProtoMessage

-0.64

Personensuche

-0.60

 autorytatywna

-0.59

 HasFactory

-0.57

 cherchés

-0.56

 lenker

-0.56

évaluateur

-0.55

 venons

-0.55

POSITIVE LOGITS

fill

0.76

 fill

0.71

 Fill

0.64

Fill

0.62

fills

0.62

 hand

0.60

 FILL

0.57

Hand

0.57

CloseOperation

0.56

 filling

0.54

Activations Density 0.467%

references to requests and formal procedures

No Comments

No Known Activations