INDEX

Explanations

instances of conditional or comparative statements reflecting on possibilities and choices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

Either

-0.87

 Either

-0.85

either

-0.81

 either

-0.80

 prostu

-0.74

 even

-0.67

 entweder

-0.67

 invece

-0.65

更是

-0.63

 instead

-0.63

POSITIVE LOGITS

 technically

1.30

 nominally

1.20

 outwardly

1.11

 ostensibly

1.11

 superfic

1.08

 theoretically

1.06

 admittedly

1.01

 initially

0.98

may

0.98

 certes

0.93

Activations Density 0.528%

instances of conditional or comparative statements reflecting on possibilities and choices

No Comments

No Known Activations