INDEX

Explanations

phrases indicating possibility or methods of achieving something

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

no way to

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

BagLayout

-0.50

期刊论文

-0.49

 HasFactory

-0.47

lgari

-0.47

dafx

-0.46

 administratif

-0.46

parsedMessage

-0.45

Portale

-0.45

 ventilador

-0.44

 iconTwitter

-0.44

POSITIVE LOGITS

 impossible

0.61

impossible

0.59

Impossible

0.58

 nowhere

0.55

 Impossible

0.53

Nowhere

0.51

 chance

0.51

 whatsoever

0.50

 Nowhere

0.50

anyone

0.48

Activations Density 0.021%

phrases indicating possibility or methods of achieving something

no way to

No Comments

No Known Activations