INDEX

Explanations

quantitative comparisons, such as measurements, values, or quantities

oai_token-act-pair · gpt-3.5-turbo

phrases related to comparisons and quantifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ffe

-0.79

rant

-0.75

NetMessage

-0.70

Ã©r

-0.67

hid

-0.67

camp

-0.64

éĥ

-0.63

rupt

-0.63

oller

-0.63

çīĪ

-0.62

POSITIVE LOGITS

 equivalent

1.10

 astronomical

1.10

 impressive

1.09

 peanuts

1.09

 hefty

1.03

 enough

1.02

 negligible

1.02

 unheard

1.00

 insane

1.00

 HUGE

1.00

Activations Density 0.181%

quantitative comparisons, such as measurements, values, or quantities

phrases related to comparisons and quantifications

No Comments

No Known Activations