INDEX

Explanations

questions and suggestions related to the word "should."

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

WAUKEE

-0.45

 приходится

-0.40

jsonwebtoken

-0.38

iseite

-0.36

égard

-0.35

wußt

-0.35

https

-0.35

reactivex

-0.35

localctx

-0.35

feri

-0.35

POSITIVE LOGITS

Should

2.91

 Should

2.84

should

2.75

 should

2.69

 SHOULD

2.20

hould

2.17

 fhould

2.05

 debería

1.95

 deberían

1.93

 devrait

1.88

Activations Density 0.200%

questions and suggestions related to the word "should."

No Known Activations