INDEX

Explanations

words associated with permanence or lasting attributes

oai_token-act-pair · gpt-4o-mini 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

Bus

-0.74

yer

-0.72

eye

-0.71

draw

-0.69

GER

-0.68

VERTIS

-0.68

 Krug

-0.68

lycer

-0.66

roads

-0.66

 Rage

-0.65

POSITIVE LOGITS

ittee

1.22

utations

1.17

ittees

1.05

utation

1.03

itic

1.00

iton

0.98

isphere

0.95

essage

0.90

itt

0.88

orrow

0.87

Activations Density 0.017%

words associated with permanence or lasting attributes

No Comments

No Known Activations