INDEX

Explanations

proper nouns related to geography, organizations, and titles

oai_token-act-pair · gpt-3.5-turbo

references to specific countries and their abbreviations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ischer

-0.77

 Horus

-0.76

epad

-0.73

 ..........

-0.72

hement

-0.71

cgi

-0.69

efully

-0.69

ochet

-0.68

nels

-0.67

anwhile

-0.67

POSITIVE LOGITS

 citiz

0.97

 citizen

0.95

 citizens

0.93

 cities

0.82

 households

0.82

 residents

0.81

 employees

0.80

 towns

0.77

sylv

0.76

 twent

0.76

Activations Density 0.237%

proper nouns related to geography, organizations, and titles

references to specific countries and their abbreviations

No Comments

No Known Activations