INDEX

Explanations

comparative phrases indicating one thing surpassing another in some way

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

 Secondary

-0.73

 Alley

-0.67

 Amph

-0.65

 Gothic

-0.63

 Tsukuyomi

-0.63

 sectarian

-0.62

 Outdoor

-0.60

usterity

-0.60

 mandatory

-0.60

clusive

-0.59

POSITIVE LOGITS

wrote

1.20

ifies

1.16

should

1.09

did

1.08

saw

1.04

iates

0.98

heartedly

0.97

pires

0.96

poses

0.94

didn

0.94

Activations Density 0.242%

comparative phrases indicating one thing surpassing another in some way

No Comments

No Known Activations