INDEX

Explanations

programming-related terms and structures

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

OrNil

-0.48

XtraGrid

-0.46

delwed

-0.40

 XCTest

-0.39

nezeu

-0.38

Gön

-0.38

DIPSETTING

-0.37

 handleMessage

-0.37

 betweenstory

-0.37

getResultList

-0.37

POSITIVE LOGITS

 FormData

0.69

 prepared

0.61

BagConstraints

0.60

 Prepared

0.55

 prepares

0.52

 preparado

0.52

 hazır

0.52

prepared

0.50

prepareStatement

0.50

 preparing

0.49

Activations Density 0.016%

programming-related terms and structures

No Comments

No Known Activations