INDEX

Explanations

phrases related to actions or events in recent history

oai_token-act-pair · gpt-3.5-turbo

events and significant occurrences in a timeline format

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

imar

-0.65

ino

-0.61

 avoids

-0.60

 denies

-0.60

 shouldn

-0.60

 knows

-0.60

ugu

-0.58

 nowadays

-0.58

 loves

-0.58

 faiths

-0.58

POSITIVE LOGITS

 again

0.84

Reviewer

0.81

USS

0.70

FTWARE

0.69

 Again

0.67

reetings

0.66

again

0.65

AGA

0.64

amorph

0.62

iphany

0.62

Activations Density 0.957%

phrases related to actions or events in recent history

events and significant occurrences in a timeline format

No Comments

No Known Activations