INDEX

Explanations

phrases that involve people and their actions or identities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

 CreateTagHelper

-0.61

addContainerGap

-0.59

rxjs

-0.59

Raycast

-0.54

 AppDelegate

-0.53

ModelSerializer

-0.52

 Protección

-0.52

┬

-0.52

 débats

-0.51

 larmes

-0.51

POSITIVE LOGITS

who

0.82

 którzy

0.72

 quienes

0.66

 ktorí

0.62

 living

0.59

 kteří

0.57

issippi

0.57

TagMode

0.56

ingtones

0.53

 אחרים

0.53

Activations Density 0.290%

phrases that involve people and their actions or identities

No Comments

No Known Activations