INDEX

Explanations

references to documents and their types within technical or bureaucratic contexts

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

 customer

-0.27

slices

-0.26

 conta

-0.25

 mission

-0.24

 front

-0.22

 flood

-0.22

 deberes

-0.22

 surface

-0.22

 retention

-0.22

 base

-0.21

POSITIVE LOGITS

 للمعارف

1.13

ValueStyle

1.01

ftagPool

0.97

 EconPapers

0.91

 يتيمه

0.89

ագրություններ

0.88

 שוליים

0.86

ConstraintMaker

0.86

fjspx

0.85

Espèce

0.85

Activations Density 0.036%

references to documents and their types within technical or bureaucratic contexts

No Comments

No Known Activations