INDEX

Explanations

references to positive and negative experiences or outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

 Portuguese

-0.79

EVA

-0.75

eur

-0.74

 Mecca

-0.71

Øª

-0.68

 Saud

-0.65

Äĩ

-0.65

uana

-0.64

Ùĩ

-0.63

Bah

-0.62

POSITIVE LOGITS

ides

1.17

cale

1.02

ourced

1.01

ourcing

1.00

ided

1.00

uits

1.00

ummer

0.99

hift

0.99

undown

0.99

aic

0.98

Activations Density 0.010%

references to positive and negative experiences or outcomes

No Comments

No Known Activations