INDEX

Explanations

phrases related to specific events or incidents, particularly those involving violence or tragedy

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

 Cosponsors

-0.82

iatric

-0.70

hiba

-0.70

appropriate

-0.69

Versions

-0.68

Choice

-0.65

PDF

-0.65

Password

-0.64

accompan

-0.64

umatic

-0.63

POSITIVE LOGITS

 Baghdad

0.96

 Bethlehem

0.86

 Brisbane

0.85

 Mosul

0.84

 Interstate

0.84

 Perth

0.83

 Lima

0.83

 Aleppo

0.81

 Damascus

0.79

 Denver

0.77

Activations Density 0.036%

phrases related to specific events or incidents, particularly those involving violence or tragedy

No Comments

No Known Activations