INDEX

Explanations

positive emotions and announcements

oai_token-act-pair · gpt-3.5-turbo

expressions of positive emotions and announcements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

aghetti

-0.71

 dunno

-0.70

VERTISEMENT

-0.70

ailability

-0.69

 trailed

-0.67

igion

-0.64

arettes

-0.60

 tended

-0.59

 assum

-0.56

essage

-0.56

POSITIVE LOGITS

to

0.83

 announce

0.79

 that

0.71

dy

0.67

bly

0.66

leased

0.66

 Cosponsors

0.65

to

0.64

that

0.63

ikers

0.62

Activations Density 0.070%

positive emotions and announcements

expressions of positive emotions and announcements

No Comments

No Known Activations