INDEX

Explanations

phrases related to requests or commands for someone to leave or stop doing something

oai_token-act-pair · gpt-3.5-turbo

commands or requests related to leaving, stopping, and staying

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 glimpse

-0.61

eria

-0.61

ixture

-0.60

amoto

-0.60

 molded

-0.59

NetMessage

-0.59

culosis

-0.59

 inval

-0.59

 overlap

-0.58

 stumble

-0.58

POSITIVE LOGITS

 ASAP

1.21

SPONSORED

0.79

 immediately

0.76

 uncond

0.74

 urgently

0.71

 incrim

0.70

 lest

0.69

 peacefully

0.67

him

0.66

 Immediately

0.65

Activations Density 0.324%

phrases related to requests or commands for someone to leave or stop doing something

commands or requests related to leaving, stopping, and staying

No Comments

No Known Activations