INDEX

Explanations

references to methods or approaches in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 pleaſure

-1.22

 Majefty

-1.19

 Monfieur

-1.18

вгений

-1.16

 ―――――

-1.14

 Theſe

-1.14

 preſent

-1.13

 Inſ

-1.12

 Reſ

-1.12

 myſelf

-1.11

POSITIVE LOGITS

way

2.80

Way

2.61

WAY

2.48

Way

2.47

way

2.42

WAY

2.15

 ways

1.92

 Ways

1.70

 WAYS

1.62

Ways

1.58

Activations Density 0.071%

references to methods or approaches in various contexts

No Comments

No Known Activations