INDEX

Explanations

words related to upgrading and improvement

oai_token-act-pair · gpt-3.5-turbo

terms related to enhancing or improving something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

bered

-0.78

 Unch

-0.75

erella

-0.73

pal

-0.73

 Caption

-0.71

fried

-0.68

tering

-0.68

LESS

-0.68

cheat

-0.65

avery

-0.64

POSITIVE LOGITS

grades

0.98

grading

0.93

 upgrades

0.90

ments

0.86

graded

0.83

 upgrade

0.83

xual

0.82

 downgrade

0.77

 depreciation

0.77

 subscriptions

0.75

Activations Density 0.020%

words related to upgrading and improvement

terms related to enhancing or improving something

No Comments

No Known Activations