INDEX

Explanations

instances relating to quantifying or referencing unspecified quantities

oai_token-act-pair · gpt-3.5-turbo

references to moments or occurrences that involve plurality or countable items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ansom

-0.71

versible

-0.69

ramid

-0.67

obin

-0.66

iggurat

-0.64

¿½

-0.64

ĸļ

-0.64

ELL

-0.63

spection

-0.63

ataka

-0.63

POSITIVE LOGITS

of

1.13

 guys

1.10

guy

1.07

 dudes

1.02

 folks

0.99

 dude

0.98

 devs

0.92

 buddies

0.90

 idiots

0.90

 people

0.88

Activations Density 0.256%

instances relating to quantifying or referencing unspecified quantities

references to moments or occurrences that involve plurality or countable items

No Comments

No Known Activations