INDEX

Explanations

keywords related to geographical locations and historical events

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

orian

-0.74

onom

-0.70

arians

-0.69

arist

-0.66

orians

-0.65

ummy

-0.65

 Lori

-0.64

arian

-0.64

oln

-0.63

ucks

-0.63

POSITIVE LOGITS

route

1.21

ways

1.15

roads

1.14

 junction

1.14

 routes

1.13

 crossings

1.12

 connecting

1.12

Route

1.11

 route

1.09

 crossing

1.06

Activations Density 5.129%

keywords related to geographical locations and historical events

No Comments

No Known Activations