INDEX

Explanations

articles or stories that have been updated, corrected, or amended

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the status or updates of a story

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

FTWARE

-0.78

seeing

-0.74

ometers

-0.70

Â¯Â¯

-0.67

undreds

-0.63

asking

-0.61

ciation

-0.61

reciation

-0.61

 phyl

-0.60

onse

-0.59

POSITIVE LOGITS

 reprinted

1.07

 edited

1.03

 amended

0.99

 updated

0.92

 condensed

0.90

 corrected

0.87

 editorial

0.86

 reproduced

0.83

 synd

0.82

 Subscribe

0.81

Activations Density 0.098%

articles or stories that have been updated, corrected, or amended

phrases indicating the status or updates of a story

No Comments

No Known Activations