INDEX

Explanations

words related to political commentary and government actions

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

 guiActiveUn

-0.78

INTER

-0.71

Event

-0.65

ãĤ¨ãĥ«

-0.64

 organisers

-0.64

Tag

-0.64

NB

-0.63

 completion

-0.62

amination

-0.62

TAG

-0.60

POSITIVE LOGITS

wide

1.11

men

1.00

manship

0.80

liness

0.79

velt

0.78

 democracy

0.75

 electing

0.74

ICAN

0.74

hood

0.73

thritis

0.72

Activations Density 0.088%

words related to political commentary and government actions

No Comments

No Known Activations