INDEX

Explanations

words related to physical location or position

oai_token-act-pair · gpt-3.5-turbo

references to ground-level concepts and measurements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Marketable

-0.79

uary

-0.72

nesday

-0.68

 Braz

-0.67

imedia

-0.65

anium

-0.63

iev

-0.62

 percept

-0.62

itely

-0.62

ires

-0.61

POSITIVE LOGITS

based

1.12

mounted

1.01

plan

0.92

borne

0.91

level

0.89

grown

0.89

building

0.87

condition

0.86

floor

0.86

side

0.84

Activations Density 0.106%

words related to physical location or position

references to ground-level concepts and measurements

No Comments

No Known Activations