INDEX

Explanations

words related to protective measures or items

oai_token-act-pair · gpt-3.5-turbo

terms related to protection or safety measures

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

ango

-0.81

wo

-0.70

 Grind

-0.70

Stars

-0.69

 Crunch

-0.68

peak

-0.67

 Markets

-0.64

Market

-0.63

zb

-0.63

 Cinema

-0.63

POSITIVE LOGITS

 protective

3.81

 Protective

2.65

 protection

1.75

protection

1.56

 preventive

1.41

 shielding

1.37

prot

1.36

 restraining

1.34

 protect

1.30

 Protection

1.30

Activations Density 0.016%

words related to protective measures or items

terms related to protection or safety measures

No Comments

No Known Activations