INDEX

Explanations

words related to physical objects and their locations in a room

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ôtel

-0.57

 تضيفلها

-0.52

OGND

-0.50

thè

-0.50

uminazione

-0.50

ften

-0.47

 Atenas

-0.46

ventude

-0.46

躇

-0.46

 vernac

-0.46

POSITIVE LOGITS

 mirror

1.36

 Mirror

1.30

 mirrors

1.28

mirror

1.20

Mirror

1.18

 Mirrors

1.15

 mirroring

1.09

Mirrors

1.08

 MIRROR

1.07

 mirrored

1.04

Activations Density 0.068%

words related to physical objects and their locations in a room

No Comments

No Known Activations