INDEX

Explanations

phrases related to personal preferences and opinions

oai_token-act-pair · gpt-3.5-turbo

expressions of liking or positive sentiment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 regretted

-0.74

Required

-0.74

soDeliveryDate

-0.71

pter

-0.70

iversal

-0.68

ufact

-0.67

ultane

-0.67

 transpired

-0.67

channelAvailability

-0.67

laim

-0.66

POSITIVE LOGITS

 idea

1.13

 smell

0.85

 gravy

0.82

 vibe

0.80

job

0.78

 flexibility

0.77

 outdoors

0.77

 guts

0.76

 simplicity

0.74

 taste

0.73

Activations Density 0.167%

phrases related to personal preferences and opinions

expressions of liking or positive sentiment

No Comments

No Known Activations