INDEX

Explanations

references to memory and recall

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

memory or remembering

np_acts-logits-general · gemini-2.0-flash

memory recall

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

Datagram

-0.41

さつ

-0.39

shadowColor

-0.38

 Perſ

-0.36

astero

-0.36

jss

-0.35

 perſ

-0.35

 faſt

-0.34

 Audiodateien

-0.34

мені

-0.34

POSITIVE LOGITS

rungsseite

0.80

 recollection

0.70

 Paglinawan

0.69

WireFormatLite

0.65

 remembers

0.64

 remembered

0.63

 remembering

0.62

remember

0.60

 memory

0.59

forgot

0.59

Activations Density 0.244%

references to memory and recall

memory or remembering

memory recall

No Comments

No Known Activations