INDEX

Explanations

keywords related to editing or revisions in a text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Edit" indicating editing actions or statuses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

milo

-0.90

ciating

-0.88

gart

-0.82

NetMessage

-0.75

ingen

-0.70

swer

-0.67

ueller

-0.66

emouth

-0.65

gaard

-0.65

ynt

-0.65

POSITIVE LOGITS

 edit

0.89

edit

0.73

 Edit

0.70

orship

0.67

 summ

0.66

iton

0.62

 Transcript

0.61

 edits

0.61

amount

0.60

âĶĢâĶĢ

0.60

Activations Density 0.010%

keywords related to editing or revisions in a text

instances of the word "Edit" indicating editing actions or statuses

No Comments

No Known Activations