INDEX

Explanations

terms related to job positions or roles, especially emphasizing a return to previous positions

oai_token-act-pair · gpt-3.5-turbo

terms related to past roles or positions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ARCH

-0.75

abbage

-0.75

otle

-0.72

asus

-0.72

ovember

-0.71

ghai

-0.70

 themselves

-0.70

ocene

-0.69

wyn

-0.69

zanne

-0.66

POSITIVE LOGITS

 employer

1.29

 selves

1.12

 colleague

1.09

 adversary

1.07

 haun

1.06

 partner

1.02

 hometown

1.02

 teammate

1.00

 predicament

1.00

 employers

1.00

Activations Density 0.142%

terms related to job positions or roles, especially emphasizing a return to previous positions

terms related to past roles or positions

No Comments

No Known Activations