INDEX

Explanations

phrases indicating presence and implications of certain features or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

ThroughAttribute

-1.13

Portail

-0.91

 ſind

-0.82

 Portail

-0.81

 Houſe

-0.79

 متعلقه

-0.75

 myſelf

-0.74

 fubject

-0.71

rostis

-0.70

]<<"

-0.70

POSITIVE LOGITS

 been

0.67

 betrekking

0.64

its

0.64

 become

0.62

 resulted

0.61

 nothing

0.60

 été

0.58

 implications

0.57

 ergeben

0.53

 BECOME

0.51

Activations Density 0.249%

phrases indicating presence and implications of certain features or conditions

No Comments

No Known Activations