INDEX

Explanations

references to previous instances or reports

oai_token-act-pair · gpt-3.5-turbo

references to prior events or data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 ware

-0.71

lua

-0.69

rage

-0.64

hots

-0.64

alker

-0.62

oppable

-0.62

acer

-0.61

aliation

-0.60

hop

-0.60

onics

-0.60

POSITIVE LOGITS

 generations

0.97

ebin

0.96

 incarn

0.91

 occupant

0.83

 editions

0.82

ĸļ

0.82

itized

0.81

 installments

0.80

 incarnation

0.79

 iterations

0.78

Activations Density 0.017%

references to previous instances or reports

references to prior events or data

No Comments

No Known Activations