INDEX

Explanations

names of people or entities

oai_token-act-pair · gpt-3.5-turbo

proper names and significant figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Reviewer

-0.44

 grate

-0.43

Pwr

-0.40

":"/

-0.40

 Lama

-0.38

rower

-0.36

Äĩ

-0.36

umbnails

-0.35

psons

-0.35

 Morty

-0.34

POSITIVE LOGITS

ratch

0.42

 eternity

0.39

dan

0.38

 standpoint

0.38

 ausp

0.37

idelity

0.37

 rooft

0.36

 contingency

0.35

 shores

0.35

 vantage

0.34

Activations Density 5.729%

names of people or entities

proper names and significant figures

No Comments

No Known Activations