INDEX

Explanations

underscore-prefixed or hyphenated terms, likely related to programming or technical language elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

TagMode

-0.77

 Betracht

-0.63

 suivante

-0.61

 âgées

-0.61

 themſelves

-0.61

 adicionais

-0.61

Efq

-0.60

 constipation

-0.59

المناصب

-0.59

<bos>

-0.59

POSITIVE LOGITS

########.

0.94

شهاد

0.77

__*/

0.61

 ویکی‌پدیا

0.60

spesies

0.59

发表于

0.59

an

0.58

0.57

vellous

0.55

خصة

0.55

Activations Density 0.155%

underscore-prefixed or hyphenated terms, likely related to programming or technical language elements

No Comments

No Known Activations