INDEX

Explanations

names of institutions or organizations and their associated positions or titles

oai_token-act-pair · gpt-3.5-turbo

citations from authoritative sources or experts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

worldly

-0.65

 juicy

-0.57

Ð

-0.57

oir

-0.55

 heater

-0.55

 patient

-0.54

Availability

-0.54

lift

-0.54

 status

-0.54

eed

-0.53

POSITIVE LOGITS

0.68

 Nanto

0.63

 Controls

0.61

Dru

0.61

Ala

0.60

 congratulated

0.60

ounces

0.60

Maz

0.59

 Calif

0.58

 echoing

0.58

Activations Density 0.216%

names of institutions or organizations and their associated positions or titles

citations from authoritative sources or experts

No Comments

No Known Activations