INDEX

Explanations

terms related to a specific medical condition, specifically focusing on instances of "micro" in the context of "microcephaly."

oai_token-act-pair · gpt-3.5-turbo

terms related to "micro" objects or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

deen

-0.83

adh

-0.80

¦

-0.80

vati

-0.75

alach

-0.71

Ain

-0.70

 Pride

-0.69

fell

-0.68

===

-0.68

 Albania

-0.67

POSITIVE LOGITS

 micro

3.60

micro

2.68

Micro

2.62

 Micro

2.45

 macro

1.69

 microscopic

1.56

 nano

1.54

mic

1.42

 mini

1.31

nan

1.30

Activations Density 0.021%

terms related to a specific medical condition, specifically focusing on instances of "micro" in the context of "microcephaly."

terms related to "micro" objects or concepts

No Comments

No Known Activations