INDEX

Explanations

phrases related to classified documents and investigations

oai_token-act-pair · gpt-3.5-turbo

terms related to official documents and associated claims

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

region

-0.75

rl

-0.74

alone

-0.72

Pyr

-0.71

aple

-0.68

ells

-0.68

erd

-0.67

place

-0.66

Interstitial

-0.65

Ĥ

-0.64

POSITIVE LOGITS

 dossier

1.43

ossier

0.91

 Ukrain

0.81

ikuman

0.81

 trove

0.78

 memos

0.76

 cache

0.73

 compiled

0.73

 dumps

0.71

ilion

0.71

Activations Density 0.007%

phrases related to classified documents and investigations

terms related to official documents and associated claims

No Comments

No Known Activations