INDEX

Explanations

numerical ordinals related to rankings, such as "1st" or "2nd"

oai_token-act-pair · gpt-3.5-turbo

references to ranking or positions denoted by the term "first."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iannopoulos

-0.68

ETS

-0.65

 therap

-0.64

MpServer

-0.63

 SPACE

-0.63

 ratios

-0.62

HOME

-0.62

Statistics

-0.62

 Canaver

-0.61

estern

-0.59

POSITIVE LOGITS

 responders

0.81

 installment

0.79

aire

0.78

 anniversary

0.78

ablishment

0.78

century

0.77

 Anniversary

0.76

Lt

0.75

 century

0.74

ioxide

0.74

Activations Density 0.030%

numerical ordinals related to rankings, such as "1st" or "2nd"

references to ranking or positions denoted by the term "first."

No Comments

No Known Activations