INDEX

Explanations

phrases related to doing things correctly or in the right way

oai_token-act-pair · gpt-3.5-turbo

phrases related to moral or ethical decision-making

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

urated

-0.71

raltar

-0.69

 arsen

-0.68

 lodged

-0.67

icia

-0.65

 vanquished

-0.62

angered

-0.61

edia

-0.61

hner

-0.61

 settled

-0.59

POSITIVE LOGITS

 thing

1.24

 chores

1.16

 things

1.06

 stunts

1.05

 tasks

1.03

job

1.00

 homework

0.99

 flips

0.96

 calculations

0.96

thing

0.94

Activations Density 0.242%

phrases related to doing things correctly or in the right way

phrases related to moral or ethical decision-making

No Comments

No Known Activations