INDEX

Explanations

mentions of specific individuals or accounts on social media platforms like Twitter

oai_token-act-pair · gpt-3.5-turbo

social media handles and references

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

ãĥĺ

-0.89

ADRA

-0.86

 Lena

-0.85

 Bradford

-0.81

 Beacon

-0.81

ahar

-0.80

 Larson

-0.80

 corrid

-0.79

 Tanz

-0.78

adra

-0.76

POSITIVE LOGITS

 Mike

1.74

Mike

1.72

 Michael

1.22

Michael

1.21

MJ

0.99

MJ

0.97

Jes

0.96

 Moose

0.92

mos

0.91

meier

0.90

Activations Density 0.331%

mentions of specific individuals or accounts on social media platforms like Twitter

social media handles and references

No Comments

No Known Activations