INDEX

Explanations

names related to political figures

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of specific individuals or entities, particularly political figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

advertisement

-0.79

pmwiki

-0.78

ISSION

-0.76

ADE

-0.76

é¾

-0.75

izations

-0.75

birth

-0.74

ä¸ī

-0.72

ords

-0.70

ESSION

-0.70

POSITIVE LOGITS

 Tayyip

1.00

 Erdogan

0.86

hammad

0.84

Erd

0.83

selage

0.83

nesday

0.81

ratulations

0.79

gey

0.78

 Ramadan

0.74

gie

0.73

Activations Density 0.004%

names related to political figures

repeated mentions of specific individuals or entities, particularly political figures

No Comments

No Known Activations