INDEX

Explanations

words related to statistical analysis or research findings

oai_token-act-pair · gpt-3.5-turbo

concepts related to the prevalence and incidence of conditions or phenomena

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

rade

-0.67

ateur

-0.67

 Beautiful

-0.63

rocket

-0.63

ather

-0.62

friends

-0.61

iner

-0.61

 Rocket

-0.61

 Boss

-0.60

emon

-0.60

POSITIVE LOGITS

 prevalence

3.27

 incidence

2.72

 preval

2.12

 severity

1.61

 onset

1.51

 susceptibility

1.43

 stigma

1.38

 efficacy

1.34

 occurrence

1.29

 likelihood

1.28

Activations Density 0.015%

words related to statistical analysis or research findings

concepts related to the prevalence and incidence of conditions or phenomena

No Comments

No Known Activations