INDEX

Explanations

mentions of monitoring activities or systems

oai_token-act-pair · gpt-3.5-turbo

terms related to observation and supervision

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

illard

-0.83

Anth

-0.80

erella

-0.76

hire

-0.76

 Masquerade

-0.74

onne

-0.74

anne

-0.74

ONT

-0.73

bery

-0.73

ALSE

-0.72

POSITIVE LOGITS

 monitoring

1.10

 monitors

1.10

 monitored

1.10

 monitor

1.04

itored

0.93

 Monitor

0.90

 closely

0.90

dog

0.82

 developments

0.82

 attent

0.79

Activations Density 0.040%

mentions of monitoring activities or systems

terms related to observation and supervision

No Comments

No Known Activations