INDEX

Explanations

recurrent patterns or loops within code structures

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

 Corinth

-0.54

 Diſ

-0.53

 intercession

-0.52

 Cicero

-0.52

 creature

-0.51

 estimés

-0.51

openg

-0.49

gro

-0.49

 regions

-0.49

 moderation

-0.49

POSITIVE LOGITS

 تضيفلها

0.90



0.82

первых

0.78

 ostavi

0.78

 typelib

0.72

])):

0.71

didSet

0.71

MENAFN

0.69

niająca

0.66

 ModelExpression

0.66

Activations Density 0.052%

recurrent patterns or loops within code structures

No Comments

No Known Activations