INDEX

Explanations

words related to mechanical parts or objects, specifically 'spokes' from bicycles or wheels

oai_token-act-pair · gpt-3.5-turbo

terms related to transportation and organizational roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 hump

-0.77

 Spielberg

-0.70

 Tarant

-0.70

herty

-0.69

oy

-0.67

 Bombay

-0.67

 Kubrick

-0.67

 Skywalker

-0.66

tered

-0.65

 Berkshire

-0.65

POSITIVE LOGITS

pace

0.94

ystem

0.93

 spokes

0.93

paces

0.91

press

0.89

leaders

0.87

leader

0.87

pire

0.86

cloth

0.83

worn

0.83

Activations Density 0.015%

words related to mechanical parts or objects, specifically 'spokes' from bicycles or wheels

terms related to transportation and organizational roles

No Comments

No Known Activations