INDEX

Explanations

negations or words indicating the absence of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 pleaſure

-0.98

 myſelf

-0.95

 fhew

-0.93

 itſelf

-0.93

 raiſ

-0.89

 themſelves

-0.86

 ſever

-0.85

 purpoſe

-0.85

 ſtate

-0.85

 himſelf

-0.84

POSITIVE LOGITS

not

1.60

Not

1.40

not

1.34

NOT

1.27

Not

1.23

NOT

1.12

isNot

0.96

 nicht

0.88

 niet

0.84

IsNot

0.79

Activations Density 0.241%

negations or words indicating the absence of something

No Comments

No Known Activations