INDEX

Explanations

information about individuals, such as their professions and roles

oai_token-act-pair · gpt-3.5-turbo

individuals identified by their professions or roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

fw

-0.70

WAY

-0.67

éĹĺ

-0.63

QB

-0.61

ERC

-0.60

onyms

-0.59

ensus

-0.58

ciation

-0.58

ANI

-0.58

ubi

-0.58

POSITIVE LOGITS

Joined

0.69

ÃĥÃĤ

0.69

 Latter

0.68

 Magnum

0.65

 aspiring

0.64

querque

0.63

©¶æ

0.60

uler

0.60

 earthqu

0.59

 Emer

0.59

Activations Density 0.271%

information about individuals, such as their professions and roles

individuals identified by their professions or roles

No Comments

No Known Activations