INDEX

Explanations

words related to relationships with coworkers

oai_token-act-pair · gpt-3.5-turbo

references to workers and their experiences

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 format

-0.72

 overt

-0.70

 offence

-0.66

 preseason

-0.63

 multiplayer

-0.61

 genre

-0.60

 zoom

-0.60

dis

-0.59

 obfusc

-0.59

 amphib

-0.59

POSITIVE LOGITS

workers

4.25

worker

3.61

working

1.79

 Workers

1.73

 Worker

1.56

work

1.47

 workers

1.42

Work

1.41

 worker

1.37

WORK

1.31

Activations Density 0.011%

words related to relationships with coworkers

references to workers and their experiences

No Comments

No Known Activations