INDEX

Explanations

terms related to behavior and its analysis or change

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.33.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 Tiefe

-0.42

 произ

-0.39

indro

-0.38

Rek

-0.38

Gazette

-0.37

gebraucht

-0.36

 ویکی‌پدی

-0.36

Higher

-0.36

zelfde

-0.35

 étrangère

-0.35

POSITIVE LOGITS

 behavior

0.61

 pattern

0.61

 behaviour

0.60

 Behaviour

0.59

 patterns

0.58

 Behavior

0.56

 behaviours

0.54

traits

0.54

 Pattern

0.53

behavior

0.52

Activations Density 0.158%

terms related to behavior and its analysis or change

No Known Activations