INDEX

Explanations

proper names with an apostrophe

oai_token-act-pair · gpt-3.5-turbo

instances of the letter 'O' followed by a single quote

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

vg

-0.79

 eclipse

-0.74

 incent

-0.68

 thrill

-0.67

 academics

-0.67

anwhile

-0.66

»Ĵ

-0.63

iae

-0.62

asers

-0.62

OPLE

-0.62

POSITIVE LOGITS

Brien

1.08

Neill

1.01

Connor

1.00

Malley

0.97

Donnell

0.94

clock

0.92

Leary

0.91

Reilly

0.85

Sullivan

0.82

Tel

0.79

Activations Density 0.015%

proper names with an apostrophe

instances of the letter 'O' followed by a single quote

No Comments

No Known Activations