INDEX

Explanations

phrases that express categorization or types of things

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

 purpoſe

-1.13

 pleaſure

-1.07

 cauſe

-0.99

 Diſ

-0.98

 myſelf

-0.95

 fevere

-0.95

 ſtate

-0.94

 quæ

-0.94

 uſe

-0.94

 chofe

-0.91

POSITIVE LOGITS

Kinda

1.46

 sorta

1.35

 kinda

1.30

 Kinda

1.29

kinda

1.22

 really

1.04

 pretty

0.85

really

0.84

 quite

0.79

 Really

0.77

Activations Density 0.075%

phrases that express categorization or types of things

No Comments

No Known Activations