INDEX

Explanations

phrases related to user interactions and control mechanisms

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

 Italijanski

-0.56

StructEnd

-0.55

الدراسه

-0.54

gridy

-0.54

 AssemblyProduct

-0.54

 estimés

-0.53

RetentionPolicy

-0.52

ModelSerializer

-0.51

sique

-0.51

RectangleBorder

-0.50

POSITIVE LOGITS

 action

1.84

 actions

1.84

 Actions

1.60

action

1.60

actions

1.53

Actions

1.51

Action

1.50

 Action

1.49

 ACTION

1.43

ACTION

1.42

Activations Density 0.685%

phrases related to user interactions and control mechanisms

No Comments

No Known Activations