INDEX

Explanations

proper names or locations containing the letters "ls"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the token "ls"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

PDATE

-0.69

ALLY

-0.63

 hemor

-0.62

EEE

-0.62

 partic

-0.61

TRUMP

-0.59

NIC

-0.59

 Galile

-0.58

 Berks

-0.57

quickShipAvailable

-0.57

POSITIVE LOGITS

creen

1.32

ayers

1.03

mas

1.02

ystem

1.01

pie

1.00

bach

0.95

worth

0.95

heed

0.95

leys

0.95

aurus

0.94

Activations Density 0.011%

proper names or locations containing the letters "ls"

occurrences of the token "ls"

No Comments

No Known Activations