INDEX

Explanations

identifiers and keys from a structured data format

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_131k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.33.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

ſelves

-0.60

ſelf

-0.57

 itſelf

-0.54

 ſol

-0.52

 pleaſure

-0.50

 uſed

-0.50

 sää

-0.49

 himſelf

-0.49

 myſelf

-0.49

 houſe

-0.48

POSITIVE LOGITS

VIAF

0.45

InitStruct

0.41

CppMethod

0.41

astify

0.41

andExpect

0.38

ViewImports

0.38

FormTagHelper

0.38

VersionUID

0.35

 صوتيه

0.35

 actualidad

0.35

Activations Density 0.063%

identifiers and keys from a structured data format

No Comments

No Known Activations