INDEX

Explanations

terms related to description and representation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 nemlig

-0.36

 âgées

-0.35

 ấn

-0.35

 acept

-0.35

 contribute

-0.35

 tạp

-0.35

 beina

-0.34

 nød

-0.34

 taget

-0.33

 venner

-0.33

POSITIVE LOGITS

 EconPapers

0.66

󠁣

0.61

UnusedPrivate

0.59

RectangleBorder

0.56

Portale

0.56

 utafitiHapana

0.52

BASEPATH

0.51

InjectAttribute

0.51

how

0.50

ttemberg

0.49

Activations Density 1.182%

terms related to description and representation

No Comments

No Known Activations