INDEX

Explanations

words related to proper nouns, especially names

oai_token-act-pair · gpt-3.5-turbo

names or titles of individuals or entities

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

åī

-0.95

 enthusi

-0.90

 Dynam

-0.81

 Azerb

-0.80

Mub

-0.77

 backlog

-0.76

 garg

-0.76

Bub

-0.76

kb

-0.76

 Rosenberg

-0.75

POSITIVE LOGITS

ce

1.23

ire

1.15

onse

1.15

te

1.15

ide

1.15

esse

1.15

asse

1.13

ote

1.12

le

1.12

alle

1.10

Activations Density 0.178%

words related to proper nouns, especially names

names or titles of individuals or entities

No Comments

No Known Activations