INDEX

Explanations

references to the quality of work or performance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.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

RectangleBorder

-0.79

 snippetHide

-0.57

TestBed

-0.55

ingdon

-0.55

MENAFN

-0.54

 jobbet

-0.53

 OFDb

-0.53

ptonshire

-0.52

Espèce

-0.52

 ägg

-0.51

POSITIVE LOGITS

 betweenstory

0.60

 resourceCulture

0.54

DataSnapshot

0.53

{?>

0.50

UIControlState

0.49

encodeWith

0.48

windowFixed

0.47

:])

0.45

enderror

0.44

ResumeLayout

0.44

Activations Density 0.015%

references to the quality of work or performance

No Comments

No Known Activations