INDEX

Explanations

phrases that indicate barriers or obstacles in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.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

揄

-0.37

baden

-0.36

BeNil

-0.34

 ujednoznacz

-0.34

MemoryWarning

-0.33

StreetMap

-0.33

MAH

-0.32

WithMany

-0.32

AndEndTag

-0.32

 unprotected

-0.31

POSITIVE LOGITS

 blocking

1.69

 preventing

1.66

 hindering

1.63

 impede

1.60

 obstructing

1.57

 inhibiting

1.54

 hinder

1.52

 prevent

1.51

 impediment

1.51

 prevents

1.50

Activations Density 1.361%

phrases that indicate barriers or obstacles in various contexts

No Comments

No Known Activations