INDEX

Explanations

terms related to political discussions and heated arguments

oai_token-act-pair · gpt-3.5-turbo

phrases related to political strategy and discourse

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

 Norn

-0.82

ãĤ´ãĥ³

-0.79

 Annotations

-0.76

byss

-0.74

ufact

-0.74

onding

-0.71

oided

-0.70

saf

-0.68

imum

-0.67

surprisingly

-0.67

POSITIVE LOGITS

Democrats

1.38

 Republicans

1.32

partisan

1.30

Republican

1.28

 partisan

1.27

 Republican

1.27

Republicans

1.27

 Democrats

1.24

Hillary

1.23

 Hillary

1.22

Activations Density 0.797%

terms related to political discussions and heated arguments

phrases related to political strategy and discourse

No Comments

No Known Activations