INDEX

Explanations

concepts and terms related to witchcraft and Halloween

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

ConstraintMaker

-0.73

extAlignment

-0.66

elemField

-0.61

 Issus

-0.59

RuleContext

-0.59

KURZBESCHREIBUNG

-0.57

 AssemblyProduct

-0.56

 iprot

-0.56

artifactId

-0.56

InputBorder

-0.56

POSITIVE LOGITS

 spooky

0.95

 Haunted

0.89

👻

0.88

 ghost

0.87

 ghostly

0.86

 Witch

0.84

spooky

0.83

 haunted

0.82

 witches

0.82

 ghosts

0.82

Activations Density 0.790%

concepts and terms related to witchcraft and Halloween

No Comments

No Known Activations