INDEX

Explanations

expressions of gratitude and acknowledgments in texts

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

thanking people for help

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

transQ

-0.43

closedir

-0.43

 bluzka

-0.43

 CURIAM

-0.42

jemahan

-0.42

 pembunuhan

-0.40

 bParam

-0.40

 HttpNotFound

-0.39

PerformLayout

-0.39

Hochspringen

-0.39

POSITIVE LOGITS

Dtor

0.43

 trustee

0.43

erview

0.40

 Trustee

0.39

 postgraduate

0.38

REW

0.37

Seat

0.37

انتهای

0.37

 Exter

0.37

DockStyle

0.37

Activations Density 0.020%

expressions of gratitude and acknowledgments in texts

thanking people for help

No Comments

No Known Activations