INDEX

Explanations

references to specific events or situations related to controversial topics or news

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

 guiActiveUnfocused

-0.81

+/-

-0.69

 pregn

-0.66

 ..."

-0.66

[|

-0.63

___

-0.58

classic

-0.58

 Voltage

-0.57

 Starg

-0.56

Temperature

-0.56

POSITIVE LOGITS

zbollah

1.69

pherd

1.41

ldon

1.30

odore

1.17

ffield

1.15

resy

1.11

miah

1.10

pard

1.08

ather

1.06

avier

1.06

Activations Density 0.116%

references to specific events or situations related to controversial topics or news

No Comments

No Known Activations