INDEX

Explanations

words related to football (soccer), including player names and team information

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

[*]

-0.66

 lodging

-0.60

CBI

-0.59

 toile

-0.57

 Notting

-0.57

blooded

-0.56

 questioning

-0.56

mony

-0.56

Balt

-0.56

 duct

-0.55

POSITIVE LOGITS

intosh

0.95

ufact

0.94

gage

0.84

nesota

0.82

arella

0.81

ICAN

0.81

ONEY

0.80

nih

0.77

enzie

0.76

udi

0.75

Activations Density 1.544%

words related to football (soccer), including player names and team information

No Comments

No Known Activations