INDEX

Explanations

emphasized adjectives that convey a strong sentiment or opinion

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_2/width_16k/average_l0_141

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 Cæsar

-0.74

 vertes

-0.72

 Monfieur

-0.72

 Breton

-0.71

Portail

-0.68

μία

-0.66

Cey

-0.66

 Flemish

-0.65

 chops

-0.64

 Valeria

-0.64

POSITIVE LOGITS

so

2.98

So

2.22

So

2.11

so

2.11

SO

1.96

så

1.62

 sooo

1.60

 sooooo

1.59

 soooo

1.56

SO

1.47

Activations Density 0.094%

emphasized adjectives that convey a strong sentiment or opinion

No Comments

No Known Activations