INDEX

Explanations

words related to high stakes situations or important decisions

oai_token-act-pair · gpt-3.5-turbo

terms related to risks, prospects, and opportunities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 hears

-0.72

votes

-0.67

gery

-0.67

SHIP

-0.64

 participates

-0.63

 conducts

-0.62

ifies

-0.61

 installs

-0.61

 Kills

-0.61

mits

-0.60

POSITIVE LOGITS

are

1.37

 aren

1.33

 vary

1.32

 weren

1.28

 differ

1.25

 were

1.21

 abound

1.14

 seem

1.03

 remain

1.02

 collide

1.01

Activations Density 0.319%

words related to high stakes situations or important decisions

terms related to risks, prospects, and opportunities

No Comments

No Known Activations