INDEX

Explanations

comparative phrases indicating higher levels or measurements

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

 للمعارف

-0.40

 unarmed

-0.38

 Nied

-0.38

 נח

-0.37

 nied

-0.36

raszamy

-0.36

 Eink

-0.36

zieher

-0.36

∊

-0.35

POSITIVE LOGITS

Bigger

0.63

 Bigger

0.61

outWeight

0.60

bigger

0.57

ValueStyle

0.56

 bigger

0.54

 higher

0.52

Higher

0.50

 Higher

0.50

 nakalista

0.49

Activations Density 1.764%

comparative phrases indicating higher levels or measurements

No Comments

No Known Activations