INDEX

Explanations

references to events or concepts related to a specific event

oai_token-act-pair · gpt-3.5-turbo

the term "post" followed by various numerical designations or contexts indicating a time period or thematic categorization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Qiao

-0.80

 Bett

-0.69

 Beans

-0.63

ï¸

-0.63

onge

-0.61

 bands

-0.61

 Brooks

-0.60

 Bras

-0.58

 Scouts

-0.57

 Ernst

-0.57

POSITIVE LOGITS

mortem

1.52

doctoral

1.42

graduate

1.38

gres

1.32

mort

1.27

cards

1.26

hum

1.23

secondary

1.23

modern

1.23

greSQL

1.20

Activations Density 0.029%

references to events or concepts related to a specific event

the term "post" followed by various numerical designations or contexts indicating a time period or thematic categorization

No Comments

No Known Activations