INDEX

Explanations

punctuation marks, specifically periods

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

starts of text segments

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

 GenerationType

-0.44

 مرئيه

-0.42

 Phry

-0.40

 SqlCommand

-0.40

sb

-0.40

">//

-0.40

 newOwner

-0.40

 ddelweddau

-0.39

 budapest

-0.39

inetic

-0.39

POSITIVE LOGITS

 <<<<<<<<<<<<<<

0.59

сылкі

0.53

httphttps

0.50

 répondu

0.46

 Wahrnehmung

0.45

interprétation

0.44

 pohled

0.42

 mogelijk

0.42

 wikipagina

0.41

aurait

0.41

Activations Density 0.000%

punctuation marks, specifically periods

starts of text segments

No Comments

No Known Activations