INDEX

Explanations

references to trains

oai_token-act-pair · gpt-3.5-turbo

mentions of various "Train" names or references in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 arts

-0.72

 board

-0.67

 donor

-0.66

 occas

-0.65

 conven

-0.65

 warming

-0.64

 livest

-0.64

 cons

-0.63

ously

-0.63

 shine

-0.63

POSITIVE LOGITS

 Train

1.04

 Tracks

0.95

ers

0.94

Track

0.91

Master

0.90

 Simulator

0.90

masters

0.89

rats

0.89

 Tickets

0.87

Train

0.85

Activations Density 0.037%

references to trains

mentions of various "Train" names or references in different contexts

No Comments

No Known Activations