INDEX

Explanations

the presence of negations or words indicating a lack or absence

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

Negative Logits

complexType

-0.43

quels

-0.40

Either

-0.39

 kasarigan

-0.38

 fortfarande

-0.38

ğraf

-0.38

不过

-0.36

quelles

-0.36

цездатний

-0.36

tanleria

-0.35

POSITIVE LOGITS

 only

0.92

 surprisingly

0.82

 necessarily

0.79

withstanding

0.77

 sure

0.77

orious

0.75

tingham

0.75

 knowing

0.75

 least

0.72

yet

0.71

Activations Density 0.248%

the presence of negations or words indicating a lack or absence

No Known Activations