INDEX

Explanations

information about person appearances in a film

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ustomed

-0.78

Ã¡n

-0.77

âĢĳ

-0.76

-0.74

obil

-0.74

bitious

-0.74

etermined

-0.73

ornings

-0.72

 ÂŃ

-0.71

ensable

-0.70

POSITIVE LOGITS

 kinda

1.08

 anyways

1.05

 shitty

1.02

lol

1.00

 devs

0.99

 stupidity

0.98

 fucking

0.95

 bullshit

0.94

 idiots

0.93

 stupid

0.93

Activations Density 1.482%

information about person appearances in a film

No Comments

No Known Activations