INDEX

Explanations

instances where the word "could" is used in a statement about potential future events or possibilities

oai_token-act-pair · gpt-3.5-turbo

conditional phrases indicating potential outcomes or possibilities

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

raining

-0.68

raint

-0.67

ving

-0.66

cloth

-0.65

 Yards

-0.62

 honors

-0.62

Maker

-0.62

teen

-0.61

hibited

-0.60

core

-0.60

POSITIVE LOGITS

 conce

1.18

 feas

1.17

be

1.03

berra

1.01

adian

0.93

 potentially

0.93

 possibly

0.92

 theoretically

0.91

 conclud

0.90

 hypot

0.90

Activations Density 0.087%

instances where the word "could" is used in a statement about potential future events or possibilities

conditional phrases indicating potential outcomes or possibilities

No Comments

No Known Activations