INDEX

Explanations

words and phrases related to spatial orientation or positioning

oai_token-act-pair · gpt-3.5-turbo

references to axes in various contexts, particularly in scientific or medical discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.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

esters

-0.93

urrent

-0.83

ãģ®éŃĶ

-0.76

namese

-0.76

ORE

-0.68

ricting

-0.67

rov

-0.67

earable

-0.66

PE

-0.66

brance

-0.66

POSITIVE LOGITS

 axis

1.37

 axes

1.02

axis

0.97

 mund

0.96

 Axis

0.89

wheel

0.82

 crank

0.82

 jerk

0.81

 slic

0.81

illary

0.78

Activations Density 0.009%

words and phrases related to spatial orientation or positioning

references to axes in various contexts, particularly in scientific or medical discussions

No Comments

No Known Activations