INDEX

Explanations

phrases that rank or evaluate performance in competitive contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

PMailer

-0.66

styleType

-0.60

 onAnimation

-0.58

AutoresizingMask

-0.57

weh

-0.57

Gön

-0.56

chaffenheit

-0.54

reht

-0.52

atypes

-0.52

 AppBundle

-0.52

POSITIVE LOGITS

 ranked

0.87

 ranking

0.81

 rankings

0.76

Ranking

0.64

 Ranking

0.63

 Ranked

0.62

top

0.61

ranked

0.59

 rank

0.59

Rankings

0.58

Activations Density 0.154%

phrases that rank or evaluate performance in competitive contexts

No Comments

No Known Activations