INDEX

Explanations

specific details or actions related to a topic

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize specific intent or detail

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 undoubtedly

-0.70

 equally

-0.68

 steadily

-0.68

 understandably

-0.67

ulton

-0.67

 predictably

-0.66

Progress

-0.65

anon

-0.65

 Isles

-0.64

 inevitably

-0.64

POSITIVE LOGITS

 targeted

1.00

 tailored

0.94

 designed

0.93

 exempted

0.89

 formulated

0.88

 targeting

0.86

atered

0.83

 addressed

0.82

 requested

0.81

designed

0.79

Activations Density 0.033%

specific details or actions related to a topic

phrases that emphasize specific intent or detail

No Comments

No Known Activations