INDEX

Explanations

statistically significant positive or negative correlations in various contexts

oai_token-act-pair · gpt-3.5-turbo

statistical associations related to health and medical research

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

olit

-0.81

aution

-0.77

amera

-0.77

adobe

-0.76

iscons

-0.72

solete

-0.72

submit

-0.72

endars

-0.71

visors

-0.71

reet

-0.71

POSITIVE LOGITS

 correlation

1.66

 correlations

1.48

 heterogeneity

1.42

 inverse

1.41

 association

1.38

 difference

1.26

 associations

1.24

 tendency

1.24

 variability

1.24

 gradient

1.23

Activations Density 0.235%

statistically significant positive or negative correlations in various contexts

statistical associations related to health and medical research

No Comments

No Known Activations