INDEX

Explanations

editor's notes in a written document

oai_token-act-pair · gpt-3.5-turbo

editorial notes or annotations within a text

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

addons

-0.67

 scrim

-0.65

ardless

-0.65

 suspic

-0.64

pload

-0.63

gur

-0.61

evil

-0.60

 dirt

-0.60

ells

-0.60

stained

-0.59

POSITIVE LOGITS

0.88

BOOK

0.84

":"","

0.83

:]

0.81

*)

0.78

>:

0.77

*:

0.76

 EDIT

0.74

::

0.73

:,

0.72

Activations Density 0.029%

editor's notes in a written document

editorial notes or annotations within a text

No Comments

No Known Activations