INDEX

Explanations

terms related to commitment or adherence to a particular course of action

oai_token-act-pair · gpt-3.5-turbo

the repeated use of the word "sticking."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

unes

-0.81

RT

-0.73

une

-0.71

BER

-0.67

uned

-0.64

uf

-0.64

OTO

-0.63

ept

-0.63

ogram

-0.62

eded

-0.62

POSITIVE LOGITS

 sticking

1.13

 plaster

0.99

 stick

0.91

 caut

0.84

 suspic

0.83

 sticks

0.82

 slic

0.81

 proble

0.78

 burner

0.77

pole

0.76

Activations Density 0.007%

terms related to commitment or adherence to a particular course of action

the repeated use of the word "sticking."

No Comments

No Known Activations

terms related to commitment or adherence to a particular course of action

the repeated use of the word "sticking."

No Comments

No Known Activations