INDEX

Explanations

words related to statistical relationships or connections between different variables

oai_token-act-pair · gpt-3.5-turbo

concepts related to correlation and causation in data

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

 Tent

-0.79

Beg

-0.68

istar

-0.68

rown

-0.65

 Asylum

-0.64

aughs

-0.64

ews

-0.64

oug

-0.64

 withdrawn

-0.63

bold

-0.62

POSITIVE LOGITS

 correlation

3.43

 correlations

2.86

 correlated

2.19

 correlate

2.14

relation

1.89

 correl

1.87

 causation

1.69

 correlates

1.65

 predictive

1.48

 predictor

1.42

Activations Density 0.029%

words related to statistical relationships or connections between different variables

concepts related to correlation and causation in data

No Comments

No Known Activations