INDEX

Explanations

the quantification of the number of instances or occurrences

oai_token-act-pair · gpt-3.5-turbo

phrases that inquire about quantities or the number of items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

acus

-0.77

FontSize

-0.66

ital

-0.66

hern

-0.66

bringing

-0.63

UAL

-0.63

roup

-0.61

rax

-0.60

Mobil

-0.60

works

-0.60

POSITIVE LOGITS

 times

0.94

een

0.84

 thousand

0.82

 dozen

0.78

 iterations

0.75

 hundred

0.74

eenth

0.73

ones

0.72

Kai

0.70

 nights

0.69

Activations Density 0.043%

the quantification of the number of instances or occurrences

phrases that inquire about quantities or the number of items

No Comments

No Known Activations