INDEX

Explanations

references to "composition" in various contexts

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

RunWith

-0.78

 tartalomajánló

-0.77

 متعلقه

-0.75

 cherchés

-0.74

ніципа

-0.74

ьаж

-0.73

Παραπομπές

-0.69

 архивлан

-0.69

alnız

-0.69

WarningLevel

-0.69

POSITIVE LOGITS

 composition

2.15

 Composition

2.12

 compositions

2.09

 compose

2.01

 composed

1.96

Composition

1.94

 compos

1.91

 Compose

1.90

composition

1.87

 COMPOSITION

1.86

Activations Density 0.072%

references to "composition" in various contexts

No Comments

No Known Activations