INDEX

Explanations

phrases related to secrecy or hidden information

oai_token-act-pair · gpt-3.5-turbo

concepts related to secrecy and confidentiality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ergy

-0.82

bal

-0.78

balance

-0.77

ermanent

-0.70

di

-0.70

arte

-0.69

imus

-0.68

opez

-0.68

Interstitial

-0.68

anza

-0.68

POSITIVE LOGITS

recy

1.14

 secrecy

1.01

 shrouded

0.92

 secrets

0.90

 shroud

0.89

 cloaked

0.84

rets

0.83

 guarded

0.78

 surrounding

0.75

 secret

0.71

Activations Density 0.026%

phrases related to secrecy or hidden information

concepts related to secrecy and confidentiality

No Comments

No Known Activations