INDEX

Explanations

numbers and financial terms

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

monetary values and references to financial metrics

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

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_post

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

 Newark

-0.69

 Rhodes

-0.68

Gad

-0.67

Osc

-0.66

 Pegasus

-0.66

 Phar

-0.63

 organisers

-0.61

 circuits

-0.60

 torpedo

-0.60

 Gent

-0.59

POSITIVE LOGITS

âĢ

1.95

 âĢ

1.71

ãĢ

1.47

âĢł

1.42

Â¨

1.23

 âĶĤ

1.22

â

1.22

â

1.18

âĺ

1.18

1.17

Activations Density 0.537%

numbers and financial terms

monetary values and references to financial metrics

No Comments

No Known Activations