INDEX

Explanations

words related to disguises or hidden identities

oai_token-act-pair · gpt-3.5-turbo

references to identity concealment and group affiliations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Ramos

-0.67

udeb

-0.66

 Cruiser

-0.65

 Bellev

-0.61

 Sherman

-0.60

 Volcano

-0.60

 Aval

-0.58

waters

-0.57

amina

-0.56

lance

-0.56

POSITIVE LOGITS

depending

1.01

 thereof

0.99

 alike

0.94

 depending

0.91

 versa

0.87

omever

0.80

atever

0.77

peat

0.77

Else

0.73

 anywhere

0.72

Activations Density 0.313%

words related to disguises or hidden identities

references to identity concealment and group affiliations

No Comments

No Known Activations