INDEX

Explanations

words related to opposition, resistance, and conflict

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĸļ

-0.71

 Pelicans

-0.70

 gorge

-0.68

 Anniversary

-0.66

 resid

-0.63

 Remastered

-0.62

 Legends

-0.62

 Gorge

-0.61

 Sparks

-0.61

Suc

-0.60

POSITIVE LOGITS

establishment

1.21

abortion

1.13

war

1.03

Semitic

0.99

measures

0.91

hero

0.91

overty

0.90

clock

0.88

immigrant

0.88

drug

0.88

Activations Density 1.428%

words related to opposition, resistance, and conflict

No Comments

No Known Activations