INDEX

Explanations

the beginning of new sentences or sections indicated by specific tokens

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 فريبيس

-0.63

جوايز

-0.55

IntoConstraints

-0.54

 Pristup

-0.51

原始内容存档

-0.50

CodedInputStream

-0.49

 esternos

-0.48

 gedrag

-0.48

migrationBuilder

-0.48

 zeitung

-0.48

POSITIVE LOGITS

Ol

0.46

0.43

Az

0.42

Dal

0.42

Ol

0.42

PD

0.41

<eos>

0.41

 Srin

0.41

St

0.40

Activations Density 0.005%

the beginning of new sentences or sections indicated by specific tokens

No Comments

No Known Activations