INDEX

Explanations

social media handles and posts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

Data Type

torch.float32

Hook Point

blocks.8.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

cised

-0.68

CFR

-0.68

heny

-0.66

 displacement

-0.63

 Dracula

-0.63

rugged

-0.63

 Quart

-0.62

enary

-0.62

 Stevens

-0.61

zik

-0.61

POSITIVE LOGITS

username

1.12

 Username

1.10

 username

1.05

 postings

0.94

pages

0.92

 chats

0.91

Reddit

0.88

 user

0.88

 feeds

0.88

 hashtag

0.88

Activations Density 3.981%

social media handles and posts

No Comments

No Known Activations