INDEX

Explanations

phrases related to two-way interactions or relationships

oai_token-act-pair · gpt-3.5-turbo

terms related to interactive relationships or exchanges

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

usters

-0.77

asts

-0.72

ummies

-0.69

oru

-0.68

irth

-0.67

 Unle

-0.66

owship

-0.66

ulner

-0.66

ares

-0.65

astical

-0.65

POSITIVE LOGITS

 street

0.87

 handshake

0.76

 mirrors

0.76

 interaction

0.74

 interference

0.72

 doors

0.70

 polarization

0.70

finding

0.70

 conversation

0.69

ECTION

0.69

Activations Density 0.035%

phrases related to two-way interactions or relationships

terms related to interactive relationships or exchanges

No Comments

No Known Activations