INDEX

Explanations

phrases related to giving or following instructions

oai_token-act-pair · gpt-3.5-turbo

commands or directives related to following guidelines or rules

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

uable

-0.69

cest

-0.66

afety

-0.64

tu

-0.63

vu

-0.62

 overpowered

-0.62

ommod

-0.62

 condos

-0.61

 unus

-0.60

mble

-0.59

POSITIVE LOGITS

 footsteps

1.37

 closely

1.04

 path

1.02

 footprints

0.96

 instructions

0.94

 route

0.92

 steps

0.92

 directions

0.91

 paths

0.89

 trajectory

0.86

Activations Density 0.209%

phrases related to giving or following instructions

commands or directives related to following guidelines or rules

No Comments

No Known Activations