INDEX

Explanations

phrases related to boards and members

oai_token-act-pair · gpt-3.5-turbo

references to boards or board members

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Hoover

-0.67

 Leone

-0.66

 Flask

-0.64

ously

-0.64

udeau

-0.62

çīĪ

-0.62

sson

-0.62

 Caesar

-0.62

icity

-0.61

 Rover

-0.60

POSITIVE LOGITS

walk

1.41

room

1.17

rooms

1.15

members

1.14

member

1.10

 members

1.01

chair

1.00

wal

0.96

 member

0.96

 trustees

0.95

Activations Density 0.029%

phrases related to boards and members

references to boards or board members

No Comments

No Known Activations