INDEX

Explanations

the word "to" in various contexts

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-mlp/layer_6/width_131k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.6.hook_mlp_out

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

expandindo

-0.65

NameInMap

-0.60

 viewDidLoad

-0.52

ThroughAttribute

-0.52

__;

-0.51

ValueStyle

-0.47

 ErrIntOverflow

-0.45

InitVars

-0.45

 Administrativna

-0.43

 initComponents

-0.43

POSITIVE LOGITS

 actes

0.45

 ſta

0.42

 abstrato

0.41

 acuer

0.41

 themſelves

0.40

 vägen

0.38

 ſur

0.37

 vín

0.36

 Keuangan

0.36

 acte

0.36

Activations Density 0.004%

the word "to" in various contexts

No Comments

No Known Activations