INDEX

Explanations

terms related to mathematical concepts, particularly linearity and regression

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

flies

-0.80

ubs

-0.74

illon

-0.74

reon

-0.74

bourg

-0.72

rament

-0.69

rieving

-0.69

UNE

-0.68

ittee

-0.68

cham

-0.67

POSITIVE LOGITS

ity

1.23

ized

1.18

ization

1.18

izable

1.12

ities

1.02

 interpol

0.97

 algebra

0.97

izes

0.96

izing

0.95

 progression

0.92

Activations Density 0.007%

terms related to mathematical concepts, particularly linearity and regression

No Comments

No Known Activations