INDEX

Explanations

terms related to ascension and elevation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_11/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

actos

-0.32

 sarung

-0.30

quares

-0.30

megen

-0.28

 data

-0.28

merci

-0.28

tanleria

-0.28

 pipelines

-0.27

 rodz

-0.27

 wrench

-0.27

POSITIVE LOGITS

 upward

1.09

 upwards

0.98

 ascent

0.96

 climbing

0.95

 elevation

0.90

Climbing

0.90

Above

0.90

 вверх

0.89

 ascension

0.89

 above

0.89

Activations Density 0.626%

terms related to ascension and elevation

No Comments

No Known Activations