INDEX

Explanations

terms related to various types of work or tasks

oai_token-act-pair · gpt-3.5-turbo

mentions of various forms of work and employment-related discussions

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

 sidx

-0.69

je

-0.63

 stunned

-0.62

Cla

-0.60

 diabetic

-0.60

 cartoon

-0.59

 experienced

-0.58

 allergic

-0.57

 horrified

-0.57

 offended

-0.57

POSITIVE LOGITS

work

4.47

works

2.36

WORK

2.24

working

1.96

worked

1.93

Work

1.81

 Work

1.66

 work

1.66

workers

1.55

worker

1.55

Activations Density 0.006%

terms related to various types of work or tasks

mentions of various forms of work and employment-related discussions

No Comments

No Known Activations