INDEX

Explanations

conversational elements and expressions of agreement or disagreement

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

 maisha

-0.39

ROZ

-0.36

 elektrick

-0.36

 personnelle

-0.35

 yoktur

-0.34

 Performer

-0.33

족

-0.33

roz

-0.33

 oyuncak

-0.32

 Decke

-0.32

POSITIVE LOGITS

SharedCtor

0.56

principalColumn

0.54

 InputDecoration

0.54

__*/

0.53

+#+#

0.52

AndEndTag

0.51

tvguidetime

0.49

RenderAtEndOf

0.48

 HasFactory

0.47

 Infórmanos

0.47

Activations Density 0.375%

conversational elements and expressions of agreement or disagreement

No Comments

No Known Activations