INDEX

Explanations

phrases related to government and policies

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

ses

-0.80

ãģĨ

-0.71

-0.67

pees

-0.63

atical

-0.61

iking

-0.61

olves

-0.61

fell

-0.60

ema

-0.60

sk

-0.59

POSITIVE LOGITS

romeda

1.21

 furthermore

1.10

rea

1.06

 therein

1.05

 although

0.99

 secondly

0.99

 moreover

0.97

ersen

0.97

 besides

0.92

yes

0.91

Activations Density 0.075%

phrases related to government and policies

No Comments

No Known Activations