INDEX

Explanations

expressions of gratitude or positive feedback

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

thank you / exclamation

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 concentración

-0.49

Rhestr

-0.48

 conducción

-0.48

ValueStyle

-0.45

rrggbb

-0.44

 detención

-0.44

 pegatina

-0.44

 cárcel

-0.43

/\.

-0.43

 bahía

-0.43

POSITIVE LOGITS

AutoScaleMode

0.48

RenderAtEndOf

0.47

 jScrollPane

0.47

PageRoute

0.46

 receipt

0.43

inflater

0.43

 messages

0.43

ViewFeatures

0.41

hdashline

0.41

 continúas

0.40

Activations Density 0.005%

expressions of gratitude or positive feedback

thank you / exclamation

No Comments

No Known Activations

expressions of gratitude or positive feedback

thank you / exclamation

No Comments

No Known Activations