INDEX

Explanations

words related to operating procedures, rules, and ethics

oai_token-act-pair · gpt-3.5-turbo

concepts related to operation, accountability, and ethical conduct

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rush

-0.70

Lost

-0.68

 Lost

-0.65

 landfall

-0.63

INGTON

-0.61

moil

-0.61

lost

-0.60

 Tears

-0.60

interstitial

-0.60

BALL

-0.60

POSITIVE LOGITS

 appropriately

1.63

 responsibly

1.57

 wisely

1.57

 correctly

1.54

 efficiently

1.50

 properly

1.47

 accurately

1.44

appropriately

1.40

 adequately

1.39

 accordingly

1.38

Activations Density 0.507%

words related to operating procedures, rules, and ethics

concepts related to operation, accountability, and ethical conduct

No Comments

No Known Activations