INDEX

Explanations

words related to homebrewing and brewing competitions

oai_token-act-pair · gpt-3.5-turbo

terms related to brewing, particularly homebrewing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 maternity

-0.72

 reinforcement

-0.72

fem

-0.68

Fem

-0.67

 shape

-0.65

 elim

-0.64

 society

-0.64

 size

-0.64

 Nicole

-0.63

 prevention

-0.63

POSITIVE LOGITS

brew

4.73

Brew

2.45

 homebrew

1.63

beer

1.58

 brewed

1.32

 Brew

1.31

 brew

1.28

Beer

1.28

BRE

1.24

 Brewing

1.24

Activations Density 0.017%

words related to homebrewing and brewing competitions

terms related to brewing, particularly homebrewing

No Comments

No Known Activations