INDEX

Explanations

technical terms related to product specifications and instructions

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

endphp

-0.35

 szint

-0.29

éndole

-0.28

 tilgjenge

-0.28

BnF

-0.26

defineProperty

-0.26

 collected

-0.26

 mengumpulkan

-0.26

endpush

-0.25

 <>",

-0.25

POSITIVE LOGITS

 EconPapers

0.64

RenderAtEndOf

0.63

 invokingState

0.60

Билгалдахарш

0.59

oredCriteria

0.57

yntaxException

0.57

isContained

0.57



0.55

fromnode

0.54

+#+#

0.53

Activations Density 0.047%

technical terms related to product specifications and instructions

No Comments

No Known Activations