INDEX

Explanations

phrases expressing personal beliefs and opinions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 14-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_14/width_16k/average_l0_84

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.14.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

 sabar

-0.73

-0.59

 abrazo

-0.57

Pautan

-0.55

 africano

-0.54

 sagde

-0.54

mogat

-0.54

lote

-0.53

 nocturno

-0.53

 xícara

-0.53

POSITIVE LOGITS

BufferException

0.71

usercontent

0.70

 merit

0.67

 believes

0.61

 estekak

0.60

 believe

0.60

featureID

0.60

 merits

0.59

 suitable

0.59

 think

0.58

Activations Density 0.371%

phrases expressing personal beliefs and opinions

No Comments

No Known Activations