INDEX

Explanations

people or groups fulfilling specified criteria or actions

oai_token-act-pair · gpt-3.5-turbo

phrases related to participation and rewards for individuals

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Harm

-0.81

 detrim

-0.77

MpServer

-0.76

quickShipAvailable

-0.75

inventoryQuantity

-0.72

 interrogation

-0.70

sein

-0.69

examination

-0.68

SPONSORED

-0.67

 Refuge

-0.65

POSITIVE LOGITS

 subscribed

1.15

 purchased

1.09

 redeemed

1.09

 pledged

1.06

 registered

1.06

 purchase

1.01

 pledge

0.96

 register

0.95

 qualify

0.95

 enrolled

0.94

Activations Density 0.177%

people or groups fulfilling specified criteria or actions

phrases related to participation and rewards for individuals

No Comments

No Known Activations