INDEX

Explanations

words related to business activities and processes

oai_token-act-pair · gpt-3.5-turbo

occurrences of numeric values or references to money

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

 eleph

-0.88

 disputed

-0.82

 displacement

-0.80

 tram

-0.79

 negoti

-0.79

 eroded

-0.78

 waged

-0.78

 feared

-0.77

 estranged

-0.77

hed

-0.76

POSITIVE LOGITS

Anyway

1.94

Advertisements

1.93

Lastly

1.67

Thank

1.65

Here

1.61

Enjoy

1.59

Below

1.59

Finally

1.59

Thanks

1.59

Firstly

1.58

Activations Density 0.387%

words related to business activities and processes

occurrences of numeric values or references to money

No Comments

No Known Activations