INDEX

Explanations

names of organizations and institutions

oai_token-act-pair · gpt-3.5-turbo

references to institutions and organizations

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

 tricks

-0.86

ãĤ¦ãĤ¹

-0.78

 brut

-0.76

 overpowered

-0.75

 grabs

-0.73

 peppers

-0.71

 babe

-0.70

 sauce

-0.70

 behaviours

-0.69

 sweets

-0.69

POSITIVE LOGITS

 University

1.16

 Municipal

1.14

 Institute

1.13

 Department

1.12

 Museum

1.07

 Chamber

1.06

 Dept

1.05

 Bureau

1.04

 Metropolitan

1.03

 Ibid

1.03

Activations Density 0.380%

names of organizations and institutions

references to institutions and organizations

No Comments

No Known Activations