INDEX

Explanations

proper nouns, especially names of individuals

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 simultane

-0.36

ãĢ

-0.36

 thereafter

-0.35

 prepar

-0.35

 progressing

-0.35

 subsequent

-0.34

ô

-0.34

 laun

-0.34

 newcom

-0.33

 sevent

-0.33

POSITIVE LOGITS

shaw

0.36

cia

0.35

quartered

0.34

DEN

0.31

pedia

0.31

iquette

0.31

pine

0.31

rary

0.31

zn

0.30

ppy

0.29

Activations Density 1.859%

proper nouns, especially names of individuals

references to specific individuals or names

No Comments

No Known Activations