INDEX

Explanations

phrases related to time, events, descriptions, and actions

oai_token-act-pair · gpt-3.5-turbo

references to specific events, conditions, or measurements relating to time and place

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

ritis

-0.80

encers

-0.79

auts

-0.78

ocate

-0.75

haar

-0.73

¶ħ

-0.73

atron

-0.71

ngth

-0.71

=-=-=-=-

-0.71

raint

-0.68

POSITIVE LOGITS

 occasions

1.16

 basis

1.16

 behalf

1.05

 footing

0.92

 occasion

0.92

 grounds

0.88

 pretext

0.87

 fronts

0.83

eve

0.81

 bandwagon

0.80

Activations Density 1.054%

phrases related to time, events, descriptions, and actions

references to specific events, conditions, or measurements relating to time and place

No Comments

No Known Activations