INDEX

Explanations

categories and classifications of various subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 37-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_37/width_16k/average_l0_124

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.37.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

 ſever

-0.49

 itſelf

-0.48

 himſelf

-0.48

bewerken

-0.47

 pleaſure

-0.47

ſelves

-0.46

 ſeveral

-0.45

ไง

-0.44

 leſs

-0.43

 onely

-0.43

POSITIVE LOGITS

 Location

0.98

 Date

0.79

 Description

0.78

Location

0.77

 Approximate

0.73

 Type

0.70

 Duration

0.69

 Status

0.69

 location

0.68

 Venue

0.66

Activations Density 0.808%

categories and classifications of various subjects

No Comments

No Known Activations