INDEX

Explanations

information about awards or contracts being granted

oai_token-act-pair · gpt-3.5-turbo

phrases related to the concept of receiving awards or contracts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ierre

-0.71

APD

-0.69

ben

-0.67

EEK

-0.66

so

-0.63

loo

-0.63

ijn

-0.61

sen

-0.60

Alz

-0.60

POSITIVE LOGITS

 awarded

1.11

 awarding

1.01

 prizes

0.98

 awards

0.89

Winner

0.81

 award

0.77

 laure

0.76

 recipients

0.74

 accol

0.74

 bestowed

0.72

Activations Density 0.010%

information about awards or contracts being granted

phrases related to the concept of receiving awards or contracts

No Comments

No Known Activations