INDEX

Explanations

terms related to leadership and authority dynamics

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

parsedMessage

-0.71

 Numerade

-0.65

EDEFAULT

-0.59

oa̍t

-0.52

دانشنامهٔ

-0.52

قایناق‌لار

-0.50

RTEX

-0.50

featureID

-0.50

 gynhyrchwyd

-0.47

hynch

-0.45

POSITIVE LOGITS

 acrí

0.43

 ماش

0.38

 niestety

0.38

 pseudos

0.36

 unfortunately

0.36

 malheureusement

0.35

 pseudo

0.35

 dubious

0.35

 artificial

0.34

 getCount

0.34

Activations Density 1.310%

terms related to leadership and authority dynamics

No Comments

No Known Activations