INDEX

Explanations

words related to obstruction, enrollment, abduction, infection, and resurrection

oai_token-act-pair · gpt-3.5-turbo

terms related to obstruction or hindrance in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

STEM

-0.73

---------

-0.67

ItemThumbnailImage

-0.66

REDACTED

-0.66

 Elves

-0.64

 Hearts

-0.64

 Rabbit

-0.63

FORE

-0.62

 Lans

-0.62

ULTS

-0.61

POSITIVE LOGITS

ing

1.70

ments

1.52

ions

1.45

1.25

ings

1.21

resses

1.15

ively

1.14

ors

1.14

orate

1.11

ers

1.10

Activations Density 0.115%

words related to obstruction, enrollment, abduction, infection, and resurrection

terms related to obstruction or hindrance in various contexts

No Comments

No Known Activations