INDEX

Explanations

phrases indicating inclusion or belonging in different groups or concepts

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate belonging or being part of a group or community

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ptive

-0.69

 relate

-0.63

terms

-0.62

 receipts

-0.61

intend

-0.61

 discharge

-0.61

ebted

-0.61

 incurred

-0.60

andy

-0.59

 loans

-0.58

POSITIVE LOGITS

 Team

0.78

 elite

0.77

Team

0.74

 Confederation

0.74

ADRA

0.71

team

0.67

 Motorsport

0.65

 delegation

0.65

 Expedition

0.65

Unit

0.65

Activations Density 0.128%

phrases indicating inclusion or belonging in different groups or concepts

phrases that indicate belonging or being part of a group or community

No Comments

No Known Activations