INDEX

Explanations

commands or instructions related to software usage

oai_token-act-pair · gpt-3.5-turbo

the verb "use" in various contexts

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

 Empires

-0.72

 Braz

-0.68

 Barg

-0.67

Ń·

-0.66

 Griff

-0.63

 Mysteries

-0.63

whel

-0.61

Heb

-0.61

stal

-0.61

 Sandwich

-0.60

POSITIVE LOGITS

fully

0.91

USE

0.89

FUL

0.88

ancies

0.87

itures

0.86

 misuse

0.84

ickr

0.83

 leverage

0.82

 aliases

0.78

ability

0.77

Activations Density 0.056%

commands or instructions related to software usage

the verb "use" in various contexts

No Comments

No Known Activations