INDEX

Explanations

phrases or words related to users or user interfaces

oai_token-act-pair · gpt-3.5-turbo

mentions of "user" and related terms in the context of user interfaces and user experience

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

hift

-0.77

 Maiden

-0.73

Lep

-0.67

utton

-0.65

forth

-0.64

 Duchess

-0.64

kowski

-0.62

uple

-0.62

illion

-0.61

uce

-0.61

POSITIVE LOGITS

agent

1.14

interface

1.12

 interface

1.12

 interfaces

1.08

Agent

1.07

 Interface

1.01

friendly

0.97

 interaction

0.91

Interface

0.89

defined

0.88

Activations Density 0.060%

phrases or words related to users or user interfaces

mentions of "user" and related terms in the context of user interfaces and user experience

No Comments

No Known Activations