INDEX

Explanations

phrases related to achieving a high rank or position, specifically standing on a podium

oai_token-act-pair · gpt-3.5-turbo

references to podiums in various contexts, particularly related to achievement or awards

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

alg

-0.79

omal

-0.76

hab

-0.72

 Wonderland

-0.71

ju

-0.68

origin

-0.67

uel

-0.66

ISTORY

-0.65

 Wass

-0.65

esthetic

-0.63

POSITIVE LOGITS

 podium

1.17

 seiz

0.87

olicy

0.86

 stalls

0.84

 showc

0.83

wip

0.80

 speeches

0.79

atform

0.79

 chairs

0.79

antry

0.78

Activations Density 0.010%

phrases related to achieving a high rank or position, specifically standing on a podium

references to podiums in various contexts, particularly related to achievement or awards

No Comments

No Known Activations