INDEX

Explanations

words indicating conditional relationships and hypothetical situations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 19-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_19/width_16k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.19.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ⓧ

-0.70

SBATCH

-0.67

twimg

-0.64

 насељу

-0.64

AlterField

-0.61

httphttps

-0.60

 مرئيه

-0.60

sizeCache

-0.59

intios

-0.59

awtextra

-0.58

POSITIVE LOGITS

SharedCtor

0.60

 Sandman

0.49

 members

0.48

 Luxor

0.47

 Kone

0.46

 dispozici

0.45

厚

0.45

Diz

0.45

 ISNI

0.45

elesai

0.44

Activations Density 0.400%

words indicating conditional relationships and hypothetical situations

No Known Activations