INDEX

Explanations

phrases related to instructions and guidance

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

bish

-0.61

itals

-0.60

secution

-0.60

itol

-0.58

olson

-0.58

buck

-0.58

 bomber

-0.57

-0.56

 martyr

-0.56

 Spectre

-0.56

POSITIVE LOGITS

 manuals

1.12

 manual

1.03

 booklet

1.02

 Manual

0.87

 instructed

0.84

 instruct

0.82

 instructions

0.82

Instruct

0.79

book

0.79

eering

0.77

Activations Density 0.052%

phrases related to instructions and guidance

No Comments

No Known Activations