INDEX

Explanations

the word "be" followed by adjectives or other descriptors

oai_token-act-pair · gpt-3.5-turbo

instances of the verb "be" and its various forms

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

 Landing

-0.75

 Selection

-0.68

 Inher

-0.65

 Deaths

-0.64

 Downs

-0.62

PsyNetMessage

-0.61

NCT

-0.61

inished

-0.60

ropolis

-0.60

regation

-0.60

POSITIVE LOGITS

 careful

1.36

 cautious

1.27

 vigilant

1.26

 mindful

1.26

 proactive

1.18

 honest

1.15

 wary

1.13

 thankful

1.07

 diligent

1.07

 generous

1.07

Activations Density 0.158%

the word "be" followed by adjectives or other descriptors

instances of the verb "be" and its various forms

No Comments

No Known Activations