INDEX

Explanations

expressions of positivity and approval

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

hadiran

-0.86

 caucus

-0.84

osoba

-0.84

 myſelf

-0.83

 Divina

-0.80

 Serif

-0.80

 Caucus

-0.80

lern

-0.78

 Heuer

-0.78

kään

-0.78

POSITIVE LOGITS

 good

1.74

good

1.69

 Good

1.68

 GOOD

1.67

Good

1.62

GOOD

1.60

 Goodwin

1.26

 Goodman

1.19

 buena

1.05

Bad

0.98

Activations Density 0.063%

expressions of positivity and approval

No Comments

No Known Activations