INDEX

Explanations

names of people

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals and their actions or roles

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Zel

-0.95

ule

-0.89

 cryptoc

-0.83

orthern

-0.83

-0.81

 Tooth

-0.81

 Phillies

-0.80

 philos

-0.80

Chip

-0.80

raph

-0.80

POSITIVE LOGITS

 Anderson

1.77

Anderson

1.71

 Evans

1.09

 Mund

1.00

 Concord

0.92

 Henderson

0.92

 Andersen

0.91

und

0.91

 Amanda

0.90

Ev

0.88

Activations Density 0.374%

names of people

references to specific individuals and their actions or roles

No Comments

No Known Activations