INDEX

Explanations

scientific or technical terms

oai_token-act-pair · gpt-3.5-turbo

keywords and sections typical of academic documents

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

miss

-0.80

office

-0.74

 hire

-0.68

hand

-0.65

soDeliveryDate

-0.62

 masters

-0.62

lich

-0.61

 shred

-0.61

 seats

-0.61

duct

-0.59

POSITIVE LOGITS

 Overview

0.87

Overview

0.80

GROUND

0.80

 References

0.79

 Height

0.78

 Purpose

0.78

 Background

0.77

 Participants

0.76

 Edit

0.76

 Approximately

0.74

Activations Density 0.085%

scientific or technical terms

keywords and sections typical of academic documents

No Comments

No Known Activations