INDEX

Explanations

phrases related to abilities or capacities, especially in relation to their potential or effectiveness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 itinéraires

-0.46

 oprot

-0.41

 sandalias

-0.41

 najważ

-0.37

 étape

-0.35

 darb

-0.34

 söyle

-0.33

 agresión

-0.32

 historias

-0.31

 Impf

-0.31

POSITIVE LOGITS

 capability

1.10

 Capability

1.05

 ability

1.05

capacity

0.98

 capacity

0.97

capability

0.97

 Ability

0.96

 Capabilities

0.96

 capabilities

0.92

 Capacity

0.92

Activations Density 0.334%

phrases related to abilities or capacities, especially in relation to their potential or effectiveness

No Comments

No Known Activations