INDEX

Explanations

phrases indicating origin or source of participants or subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_16k/average_l0_82

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

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

is

-0.55

 مرئيه

-0.52

ľa

-0.47

To

-0.44

SuppressLint

-0.44

She

-0.44

 isEqualToString

-0.44

 nakalista

-0.43

me

-0.43

là

-0.43

POSITIVE LOGITS

Portail

0.95

sizeCache

0.94

 kteří

0.81

новништво

0.72

 zufolge

0.72

بوابة

0.71

 whom

0.71

ۜ

0.70

 ويكيپيديا

0.68

 ktorí

0.68

Activations Density 0.386%

phrases indicating origin or source of participants or subjects

No Comments

No Known Activations