INDEX

Explanations

occurrences of the dollar sign and rejected entries in data or code contexts

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

Preceding "rejected"

np_acts-logits-general · gemini-2.0-flash

rejected numerical data

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Eighth

-0.66

 Ninth

-0.62

-0.60

 Eight

-0.54

 odkazy

-0.53

 EIGHT

-0.53

 Eigh

-0.51

Personensuche

-0.51

 eighth

-0.51

POSITIVE LOGITS

1.15

 fifth

0.98

 Fifth

0.95

0.91

 five

0.91

0.90

Fifth

0.84

 cinque

0.84

⁵

0.83

 quinto

0.80

Activations Density 0.417%

occurrences of the dollar sign and rejected entries in data or code contexts

Preceding "rejected"

rejected numerical data

No Comments

No Known Activations