INDEX

Explanations

phrases describing products or services and their attributes

oai_token-act-pair · gpt-3.5-turbo

phrases indicating accompanying features or benefits

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

externalActionCode

-0.71

jab

-0.69

chan

-0.69

 Archdemon

-0.63

pmwiki

-0.63

 Trance

-0.61

Attempts

-0.60

axis

-0.60

marg

-0.58

antis

-0.58

POSITIVE LOGITS

 regards

0.94

 regard

0.92

stood

0.91

drawn

0.83

 luggage

0.75

 groceries

0.75

 respect

0.74

 plenty

0.74

 loads

0.72

 dignity

0.71

Activations Density 0.051%

phrases describing products or services and their attributes

phrases indicating accompanying features or benefits

No Comments

No Known Activations