INDEX

Explanations

phrases that express certainty or strong recommendations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_4/width_65k/average_l0_66

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.4.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

 linkovi

-1.02

GEBURTSDATUM

-0.88

 ostavi

-0.88

ьаж

-0.87

tagHelperRunner

-0.85

TypedDataSet

-0.85

 NSCoder

-0.79

estacks

-0.79

 surla

-0.78

SequentialGroup

-0.77

POSITIVE LOGITS

0.47

 palavra

0.45

”

0.44

 torno

0.43

رشف

0.43

 complex

0.42

’

0.42

do

0.42

How

0.41

And

0.41

Activations Density 0.007%

phrases that express certainty or strong recommendations

No Comments

No Known Activations