INDEX

Explanations

phrases related to physical positions and conditions

oai_token-act-pair · gpt-3.5-turbo

words related to confinement or enclosure

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Reasons

-0.65

aird

-0.64

-0.63

 Terms

-0.62

 Passage

-0.62

Money

-0.61

MENT

-0.59

é¾įå

-0.59

 Mandatory

-0.58

MENTS

-0.57

POSITIVE LOGITS

 neatly

1.27

 snug

1.22

 atop

1.17

 onto

1.10

 upright

1.10

 securely

1.09

 tightly

1.09

 firmly

1.08

 together

1.07

 upside

1.02

Activations Density 0.158%

phrases related to physical positions and conditions

words related to confinement or enclosure

No Comments

No Known Activations