INDEX

Explanations

phrases or terms related to legal matters and regulations that would be subject to scrutiny or review

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

 latt

-0.60

yip

-0.60

Zur

-0.59

 Zimmer

-0.58

 Sutherland

-0.58

 cane

-0.58

ttp

-0.56

 Euras

-0.56

kr

-0.55

Tib

-0.55

POSITIVE LOGITS

ively

1.27

ivity

1.18

ivist

1.02

ioned

1.00

ivating

1.00

ive

0.97

ivities

0.96

itatively

0.94

iving

0.93

Reviewer

0.93

Activations Density 0.015%

phrases or terms related to legal matters and regulations that would be subject to scrutiny or review

No Comments

No Known Activations