INDEX

Explanations

mentions of being in top positions in a competition or event

oai_token-act-pair · gpt-3.5-turbo

references to podiums and related award or competitive contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

-0.74

cise

-0.71

hiro

-0.70

ju

-0.70

went

-0.68

omal

-0.67

esthetic

-0.66

versely

-0.66

laus

-0.66

~~~~~~~~~~~~~~~~

-0.66

POSITIVE LOGITS

 podium

1.28

 seiz

0.97

olicy

0.86

 tremend

0.86

 pedest

0.83

 showc

0.82

 stalls

0.81

ilion

0.77

OTUS

0.76

 lect

0.74

Activations Density 0.005%

mentions of being in top positions in a competition or event

references to podiums and related award or competitive contexts

No Comments

No Known Activations