INDEX

Explanations

terms related to relationships or connections between people

oai_token-act-pair · gpt-3.5-turbo

references to relationships and social connections among characters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ourcing

-0.90

etting

-0.75

¥ŀ

-0.75

pire

-0.68

outheast

-0.68

ply

-0.67

upuncture

-0.67

aber

-0.66

itting

-0.66

hooting

-0.66

POSITIVE LOGITS

mates

0.76

friend

0.74

 counterparts

0.74

 peers

0.72

hood

0.71

 buddies

0.70

NAME

0.67

 Jaime

0.65

enary

0.65

Jess

0.65

Activations Density 0.151%

terms related to relationships or connections between people

references to relationships and social connections among characters

No Comments

No Known Activations