INDEX

Explanations

the word "unless"

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 unless

-2.52

unless

-2.38

Unless

-2.16

 Unless

-2.16

除非

-1.51

 jeżeli

-0.79

 kecuali

-0.77

FormTagHelper

-0.72

assuming

-0.71

 если

-0.70

POSITIVE LOGITS

thood

0.62

følgelig

0.58

 mergeFrom

0.57

VideoCapture

0.57

️⃣

0.53

unicaciones

0.50

qas

0.48

ooga

0.48

olz

0.47

aree

0.47

Activations Density 0.116%

the word "unless"

No Comments

No Known Activations