INDEX

Explanations

abbreviations or acronyms related to various organizations or entities

oai_token-act-pair · gpt-3.5-turbo

terms related to intensive care (IC) units

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 marqu

-0.80

waukee

-0.74

 Shepard

-0.69

vironment

-0.68

leck

-0.68

 Marqu

-0.66

aru

-0.64

 Verse

-0.63

backs

-0.63

rÃ©

-0.62

POSITIVE LOGITS

IC

3.93

NIC

1.66

IC

1.57

EC

1.46

ic

1.34

 ASIC

1.32

Ic

1.29

DF

1.26

IB

1.26

ICC

1.23

Activations Density 0.019%

abbreviations or acronyms related to various organizations or entities

terms related to intensive care (IC) units

No Comments

No Known Activations