INDEX

Explanations

information about podcasts and episodes

oai_token-act-pair · gpt-3.5-turbo

references to podcasts and episodes

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

metics

-0.81

 shading

-0.72

 camouflage

-0.71

naissance

-0.68

yang

-0.67

urga

-0.67

actionDate

-0.67

 roses

-0.66

 Lauder

-0.66

 Ernst

-0.65

POSITIVE LOGITS

 podcasts

1.77

 podcast

1.68

odcast

1.66

podcast

1.65

 Podcast

1.61

 listeners

1.58

 iTunes

1.46

 listener

1.40

Episode

1.38

 episodes

1.28

Activations Density 0.261%

information about podcasts and episodes

references to podcasts and episodes

No Comments

No Known Activations