INDEX

Explanations

references to specific line numbers in text or coding 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

 Monfieur

-1.33

Datuak

-1.29

 myſelf

-1.21

 juſt

-1.16

bibinfo

-1.14

 himſelf

-1.13

 pleaſure

-1.13

 deſt

-1.08

 ſever

-1.08

 Majefty

-1.07

POSITIVE LOGITS

 line

2.14

 Line

2.13

 LINE

2.04

 lines

2.03

Line

1.96

line

1.90

 Lines

1.84

Lines

1.67

LINE

1.66

lines

1.65

Activations Density 0.120%

references to specific line numbers in text or coding contexts

No Comments

No Known Activations

references to specific line numbers in text or coding contexts

No Comments

No Known Activations