INDEX

Explanations

numerical values that are related to averages or medians

oai_token-act-pair · gpt-3.5-turbo

terms related to averages and statistical metrics

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

 vast

-0.79

 express

-0.79

 transition

-0.77

 federation

-0.77

 spirit

-0.75

 revelation

-0.74

 decor

-0.74

 identity

-0.74

 bulk

-0.73

 purpose

-0.72

POSITIVE LOGITS

 Average

2.70

Average

1.83

 Percent

1.79

 Percentage

1.73

 Rate

1.67

 Maximum

1.60

 Minimum

1.59

 Median

1.49

 Estimated

1.49

 Amount

1.46

Activations Density 0.057%

numerical values that are related to averages or medians

terms related to averages and statistical metrics

No Comments

No Known Activations