INDEX

Explanations

proper nouns and specific names mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

references to personal identification or names

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rush

-0.67

 recons

-0.66

storms

-0.66

 brink

-0.65

Decre

-0.65

lux

-0.64

Conclusion

-0.63

 Tend

-0.62

ndum

-0.60

 contemplation

-0.60

POSITIVE LOGITS

 name

2.13

 surname

1.90

 names

1.80

 initials

1.80

name

1.78

 username

1.72

 nickname

1.69

 NAME

1.67

Name

1.64

 Name

1.63

Activations Density 0.829%

proper nouns and specific names mentioned in the text

references to personal identification or names

No Comments

No Known Activations