INDEX

Explanations

words related to specific issues or topics being discussed

oai_token-act-pair · gpt-3.5-turbo

phrases that involve discussions or debates about various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

OGR

-0.82

Laughs

-0.81

COMPLE

-0.78

à¼

-0.75

Catalog

-0.74

rites

-0.74

aughs

-0.74

quist

-0.71

sbm

-0.71

\/\/

-0.70

POSITIVE LOGITS

 whether

1.33

 fairness

1.19

 legality

1.17

 privacy

1.06

 affordability

1.04

 transparency

1.03

 sexuality

1.01

 eligibility

0.99

 accountability

0.99

 authenticity

0.99

Activations Density 0.206%

words related to specific issues or topics being discussed

phrases that involve discussions or debates about various topics

No Comments

No Known Activations