INDEX

Explanations

keywords related to rates, levels, and quantities

oai_token-act-pair · gpt-3.5-turbo

words related to ratings or evaluations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 explor

-0.70

AMI

-0.65

amar

-0.64

 carbohyd

-0.63

 recre

-0.63

 appropri

-0.62

 coron

-0.62

estern

-0.61

 sprung

-0.61

 indist

-0.60

POSITIVE LOGITS

rate

1.08

limit

0.98

rates

0.91

Rate

0.88

âķĲâķĲ

0.85

rating

0.80

hyde

0.76

 Rate

0.76

chance

0.75

rent

0.73

Activations Density 0.016%

keywords related to rates, levels, and quantities

words related to ratings or evaluations

No Comments

No Known Activations